There was much talk during the recent presidential election campaign about “shy” Trump supporters. Scott Adams theorized about them on September 9, 2016:
It’s hard to count people who are intentionally hiding [a preference for Trump]. But just for fun, let’s see if we can deduce how many so-called Shy Trump Supporters are out there.
For starters, we can say with certainty that they exist. I have a better ear for that than most of you because of my Trump blogging and my public endorsement of Clinton for my personal safety. People feel comfortable telling me privately, and also anonymously online, that they hide their Trump support from their spouse and coworkers. So we know they exist. We just don’t know how many.
We know that sometimes robocall surveys and online surveys show more Trump support than human-to-human polling. So that might be an indicator, but we don’t know what other variables are in play.
In a recent Reuters poll, 7% of respondents “refused” to vote for either Trump or Clinton. I’m guessing some Shy Trump Supporters “park” their votes with Gary Johnson (polling at 9.3%) or Jill Stein (polling at 3.3%).
But I wonder if the Shy Trump supporters are mostly parked with Johnson because of gender (consciously or unconsciously), whereas Stein is more of a real protest vote against Clinton. Anecdotally, Shy Trump Supporters tell me they do park their pre-vote preferences with Johnson. So far, none have told me they are parking their vote with Stein. (This is anecdotal, and a small sample of perhaps a dozen.)
Then you also have the question of turnout. Trump is clearly generating the most enthusiasm in public appearances. I would think that translates into more new voters.
Most of my predictions so far this election cycle have been based on what I call the Master Persuader Hypothesis. I’ll depart from that model for this prediction because this one is based on a gut feel – which I understand in my rational mind to feel identical to confirmation bias. Therefore, you should place zero confidence in my prediction.
I predict that 3% of voters are Shy Trump Supporters. As polls continue to tighten, especially in battleground states, that will be enough for an electoral landslide for Trump.
And it was an electoral landslide, popular-vote totals to the contrary notwithstanding. See my analysis in “H.L. Mencken’s Final Legacy.”
In a commentary about Adams’s post, “The ‘Shy Trump Supporter’ Hypothesis” (September 10, 2016), I said this:
So I believe that Scott Adams is right. A lot of “shy Trump supporters” are claiming that they’ll vote for Johnson, but most of them will vote — if they do vote — for Trump. My evidence? Trump’s standing in Rasmussen’s poll is strongly (r-squared = 0.6) and negatively correlated with Johnson’s standing. As voters decide that they aren’t going to waste votes on Johnson, they’ll turn (mainly) to Trump….
If Johnson’s popular-vote share slips from its current 9 percent to 3 percent on election day — which is 3 times better than his showing in 2012 — Trump would pick up 3 percentage points. On the other hand, if Stein’s support slips from its current 2 percent to 1 percent on election day — 3 times better than her showing in 2012 — Clinton would pick up 0.7 percentage point. So far, so good, for Trump.
My forecasts of Johnson’s and Stein’s slippage were just about on the money. But I was nevertheless pessimistic about Trump’s chances:
[A]s the “other-undecided” vote shrinks from its present level of 7 percent to 1 percent (a bit higher than in recent elections), Clinton will pick up 5.5 percentage points while Trump picks up only 1.3 percentage point.
Adding it up, there’s a likely gain for Trump of 4+ percentage points and a likely gain for Clinton of 6+ percentage points. Adding those numbers to Rasmussen’s latest results for Trump (39 percent) and Clinton (43 percent) yields something like 43 or 44 percent for Trump and 49 or 50 percent for Clinton.
And I was wrong. As the outcome of the recent election attests, there are a lot of shy Republicans lurking in the ranks of nominally unaffiliated voters.
How many? I have estimated their strength by analyzing the Gallup poll of party affiliations. For 303 polls conducted from January 2004 to January 2017, here are the relationships between the non-aligned respondents and those who claimed a Democrat or Republican affiliation. The non-aligned respondents are those who claimed to be independents plus the small fraction of other respondents not claiming to be Democrat, Republican, or independent:
There’s a somewhat stronger, mirror-image relationship for Republicans:
As the size of the non-aligned block shrinks, more members of that bloc choose the Republican label than choose the Democrat label. For example, referring to the first graph, a drop in the fraction claiming independent/other status from the maximum of 49 percent to the minimum of 28 percent results in a gain for Democrats of 8 percentage points. Referring to the second graph, the same reduction in the independent/other fraction results in a gain for Republicans of 13 percentage points.
The equations represent long-run averages, of course, and Democrats have had their share of success among nominally unaligned voters, especially around the time of Obama’s win in 2008. The underestimation of latent support for Democrats at that time shows up in the following graph, which represents the actual results of the 303 Gallup polls and the results derived from the equations in the first two graphs:
The key to the outcome of the 2016 election is the net change in voters claiming GOP and Democrat affiliation since the high-water mark of those claiming independent-other status, which occurred in December 2013. From then until November 2016, the percentage of voters claiming independent-other status went from 48 to 41, a drop of 7 percentage points. Of the 7 points , the Republican Party added 4.7 while the Democrat Party added 2.3. As a result, Clinton’s long-expected national landslide became a hollow, California-based, popular-vote victory — and a comfortable electoral-vote win for Trump.
Web pages that link to this post usually consist of a discussion thread whose participants’ views of the post vary from “I told you so” to “that doesn’t square with me/my experience” or “MBTI is all wet because…”. Those who take the former position tend to be persons of above-average intelligence whose MBTI types correlate well with high intelligence. Those who take the latter two positions tend to be persons who are defensive about their personality types, which do not correlate well with high intelligence. Such persons should take a deep breath and remember that high intelligence (of the abstract-reasoning-book-learning kind measured by IQ tests) is widely distributed throughout the population. As I say below, ” I am not claiming that a small subset of MBTI types accounts for all high-IQ persons, nor am I claiming that a small subset of MBTI types is populated entirely by high-IQ persons.” All I am saying is that the bits of evidence which I have compiled suggest that high intelligence is more likely — but far from exclusively — to be found among persons with certain MBTI types.
The correlations between intelligence, political leanings, and happiness are admittedly more tenuous. But they are plausible.
Leftists who proclaim themselves to be more intelligent than persons of the right do so, in my observation, as a way of reassuring themselves of the superiority of their views. They have no legitimate basis for claiming that the ranks of highly intelligent persons are dominated by the left. Leftist “intellectuals” in academia, journalism, the “arts”, and other traditional haunts of leftism are prominent because they are vocal. But they comprise a small minority of the population and should not be mistaken for typical leftists, who seem mainly to populate the ranks of the civil service, labor unions, public-school “educators”, and the unemployed. (It is worth noting that public-school teachers, on the whole, are notoriously dumber than most other college graduates.)
Again, I am talking about general relationships, to which there are many exceptions. If you happen to be an exception, don’t take this post personally. You’re probably an exceptional person.
IQ AND PERSONALITY
Some years ago I found statistics about the personality traits of high-IQ persons (those who are in the top 2 percent of the population).* The statistics pertain to a widely used personality test called the Myers-Briggs Type Indicator (MBTI), which I have taken twice. In the MBTI there are four pairs of complementary personality traits, called preferences: Extraverted/Introverted, Sensing/iNtuitive, Thinking/Feeling, and Judging/Perceiving. Thus, there are 16 possible personality types in the MBTI: ESTJ, ENTJ, ESFJ, ESFP, and so on. (For an introduction to MBTI, summaries of types, criticisms of MBTI, and links to other sources, see this article at Wikipedia. A straightforward description of the theory of MBTI and the personality traits can be found here. Detailed descriptions of the 16 types are given here.)
In summary, here is what the statistics indicate about the correlation between personality traits and IQ:
- Other personality traits being the same, an iNtuitive person (one who grasps patterns and seeks possibilities) is 25 times more likely to have a high IQ than a Sensing person (one who focuses on sensory details and the here-and-now).
- Again, other traits being the same, an Introverted person is 2.6 times more likely to have a high IQ than one who is Extraverted; a Thinking (logic-oriented) person is 4.5 times more likely to have a high IQ than a Feeling (people-oriented) person; and a Judging person (one who seeks closure) is 1.6 times as likely to have a high IQ than a Perceiving person (one who likes to keep his options open).
- Moreover, if you encounter an INTJ, there is a 22% probability that his IQ places him in the top 2 percent of the population. (Disclosure: I am an INTJ.) Next are INTP, at 14%; ENTJ, 8%; ENTP, 5%; and INFJ, 5%. (The next highest type is the INFP at 3%.) The five types (INTJ, INTP, ENTJ, ENTP, and INFJ) account for 78% of the high-IQ population but only 15% of the total population.**
- Four of the five most-intelligent types are NTs, as one would expect, given the probabilities cited above. Those same probabilities lead to the dominance of INTJs and INTPs, which account for 49% of the Mensa membership but only 5% of the general population.**
- Persons with the S preference bring up the rear, when it comes to taking IQ tests.**
A person who read an earlier version of this post claims that “one would expect to see the whole spectrum of intelligences within each personality type.” Well, one does see just that, but high intelligence is skewed toward the five types listed above. I am not claiming that a small subset of MBTI types accounts for all high-IQ persons, nor am I claiming that a small subset of MBTI types is populated entirely by high-IQ persons.
I acknowledge reservations about MBTI, such as those discussed in the Wikipedia article. An inherent shortcoming of psychological tests (as opposed to intelligence tests) is that they rely on subjective responses (e.g., my favorite color might be black today and blue tomorrow). But I do not accept this criticism:
[S]ome researchers expected that scores would show a bimodal distribution with peaks near the ends of the scales, but found that scores on the individual subscales were actually distributed in a centrally peaked manner similar to a normal distribution. A cut-off exists at the center of the subscale such that a score on one side is classified as one type, and a score on the other side as the opposite type. This fails to support the concept of type: the norm is for people to lie near the middle of the subscale.
Why was it expected that scores on a subscale (E/I, S/N, T/F, J/P) would show a bimodal distribution? How often does one encounter a person who is at the extreme end of any subscale? Not often, I wager, except in places where such extremes are likely to be clustered (e.g., Extraverts in politics, Introverts in monasteries). The cut-off at the center of each subscale is arbitrary; it simply affords a shorthand characterization of a person’s dominant traits. But anyone who takes an MBTI (or equivalent instrument) is given his scores on each of the subscales, so that he knows the strength (or weakness) of his tendencies.
Regarding other points of criticism: It is possible, of course, that a person who is familiar with MBTI tends to see in others the characteristics of their known MBTI types (i.e., confirmation bias). But has that tendency been confirmed by rigorous testing? Such testing would examine the contrary case, that is, the ability of a person to predict the type of a person whom he knows well (e.g., a co-worker or relative).
The supposed vagueness of the descriptions of the 16 types arises from the complexity of human personality; but there are differences among the descriptions, just as there are differences among individuals. According to a footnote to an earlier version of the Wikipedia article about MBTI, half of the persons who take the MBTI are able to guess their types before taking it. Does that invalidate MBTI or does it point to a more likely phenomenon, namely, that introspection is a personality-related trait, one that is more common among Introverts than Extraverts? A good MBTI instrument cuts through self-deception and self-flattery by asking the same set of questions in many different ways, and in ways that do not make any particular answer seem like the “right” one.
IQ AND POLITICS
It is hard to find clear, concise analyses of the relationship between IQ and political leanings. I offer the following in evidence that very high-IQ individuals lean strongly toward libertarian positions.
The Triple Nine Society (TNS) limits its membership to persons with IQs in the top 0.1% of the population. In an undated survey (probably conducted in 2000, given the questions about the perceived intelligence of certain presidential candidates), members of TNS gave their views on several topics (in addition to speculating about the candidates’ intelligence): subsidies, taxation, civil regulation, business regulation, health care, regulation of genetic engineering, data privacy, death penalty, and use of military force.
The results speak for themselves. Those members of TNS who took the survey clearly have strong (if not unanimous) libertarian leanings.
THE RIGHT IS SMARTER THAN THE LEFT
I count libertarians as part of the right because libertarians’ anti-statist views are aligned with the views of the traditional (small-government) conservatives who are usually Republicans. Having said that, the results reported in “IQ and Politics” lead me to suspect that the right is smarter than the left, left-wing propaganda to the contrary notwithstanding. There is additional evidence for my view.
A site called Personality Page offers some data about personality type and political affiliation. The sample is not representative of the population as a whole; the average age of respondents is 25, and introverted personalities are over-represented (as you might expect for a test that is apparently self-administered through a website). On the other hand, the results are probably unbiased with respect to intelligence because the data about personality type were not collected as part of a study that attempts to relate political views and intelligence, and there is nothing on the site to indicate a left-wing bias. (Psychologists, who tend toward leftism, have a knack for making conservatives look bad, as discussed here, here, and here. If there is a strong association between political views and intelligence, it is found among so-called intellectuals, where the herd mentality reigns supreme.)
The data provided by Personality Page are based on the responses of 1,222 individuals who took a 60-question personality test that determined their MBTI types (see “IQ and Personality”). The test takers were asked to state their political preferences, given these choices: Democrat, Republican, middle of the road, liberal, conservative, libertarian, not political, and other. Political self-labelling is an exercise in subjectivity. Nevertheless, individuals who call themselves Democrats or liberals (the left) are almost certainly distinct, politically, from individuals who call themselves Republicans, conservatives, or libertarians (the right).
Now, to the money question: Given the distribution of personality types on the left and right, which distribution is more likely to produce members of Mensa? The answer: Those who self-identify as persons of the right are 15 percent more likely to qualify for membership in Mensa than those who self-identify as persons of the left. This result is plausible because it is consistent with the pronounced anti-government tendencies of the very-high-IQ members of the Triple Nine Society (see “IQ and Politics”).
REPUBLICANS (AND LIBERTARIANS) ARE HAPPIER THAN DEMOCRATS
Some 45% of all Republicans report being very happy, compared with just 30% of Democrats and 29% of independents. This finding has also been around a long time; Republicans have been happier than Democrats every year since the General Social Survey began taking its measurements in 1972….
Of course, there’s a more obvious explanation for the Republicans’ happiness edge. Republicans tend to have more money than Democrats, and — as we’ve already discovered — people who have more money tend to be happier.
But even this explanation only goes so far. If one controls for household income, Republicans still hold a significant edge: that is, poor Republicans are happier than poor Democrats; middle-income Republicans are happier than middle-income Democrats, and rich Republicans are happier than rich Democrats.
Gallup adds this:
Republicans are significantly more likely to report excellent mental health than are independents or Democrats among those making less than $50,000 a year, and among those making at least $50,000 a year. Republicans are also more likely than independents and Democrats to report excellent mental health within all four categories of educational attainment.
There is a lot more in both sources. Read them for yourself.
Why would Republicans be happier than Democrats? Here’s my thought, Republicans tend to be conservative or libertarian (at least with respect to minimizing government’s role in economic affairs). Consider Thomas Sowell’s A Conflict of Visions:
He posits two opposing visions: the unconstrained vision (I would call it the idealistic vision) and the constrained vision (which I would call the realistic vision). As Sowell explains, at the end of chapter 2:
The dichotomy between constrained and unconstrained visions is based on whether or not inherent limitations of man are among the key elements included in each vision…. These different ways of conceiving man and the world lead not merely to different conclusions but to sharply divergent, often diametrically opposed, conclusions on issues ranging from justice to war.
Idealists (“liberals”) are bound to be less happy than realists (conservatives and libertarians) because idealists’ expectations about human accomplishments (aided by government) are higher than those of realists, and so idealists are doomed to disappointment.
All of this is consistent with findings reported by law professor James Lindgren:
[C]ompared to anti-redistributionists, strong redistributionists have about two to three times higher odds of reporting that in the prior seven days they were angry, mad at someone, outraged, sad, lonely, and had trouble shaking the blues. Similarly, anti-redistributionists had about two to four times higher odds of reporting being happy or at ease. Not only do redistributionists report more anger, but they report that their anger lasts longer. When asked about the last time they were angry, strong redistributionists were more than twice as likely as strong opponents of leveling to admit that they responded to their anger by plotting revenge. Last, both redistributionists and anti-capitalists expressed lower overall happiness, less happy marriages, and lower satisfaction with their financial situations and with their jobs or housework. [Northwestern Law and Economics Research Paper 06-29, “What Drives Views on Government Redistribution and Anti-Capitalism: Envy or a Desire for Social Dominance?”, March 15, 2011]
THE BOTTOM LINE
If you are very intelligent — with an IQ that puts you in the top 2 percent of the population — you are most likely to be an INTJ, INTP, ENTJ, ENTP, or INFJ, in that order. Your politics will lean heavily toward libertarianism or small-government conservatism. You probably vote Republican most of the time because, even if you are not a card-carrying Republican, you are a staunch anti-Democrat. And you are a happy person because your expectations are not constantly defeated by reality.
Related reading (listed chronologically):
Jeff Allen, “Conservatives: The Smartest (and Happiest) People in the Room“, Barbed Wire, February 20, 2014
James Thompson, “Election Special: Are Republicans Smarter than Democrats?”, The Unz Review, November 3, 2016
Dennis Prager, “Liberals and Conservatives are Unhappy for Different Reasons“, Townhall, February 13, 2018
John J. Ray, “Leftists Are Born Unhappy“, Dissecting Leftism, February 14, 2018
* I apologize for not having documented the source of the statistics that I cite here. I dimly recall finding them on or via the website of American Mensa, but I am not certain of that. And I can no longer find the source by searching the web. I did transcribe the statistics to a spreadsheet, which I still have. So, the numbers are real, even if their source is now lost to me.
** Estimates of the distribution of MBTI types in the U.S. population are given in two tables on page 4 of “Estimated Frequencies of the Types in the United States Population”, published by the Center for Applications of Psychological Type. One table gives estimates of the distribution of the population by preference (E, I, N, S, etc.). The other table give estimates of the distribution of the population among all 16 MBTI types. The statistics for members of Mensa were broken down by preferences, not by types; therefore I had to use the values for preferences to estimate the frequencies of the 16 types among members of Mensa. For consistency, I used the distribution of the preferences among the U.S. population to estimate the frequencies of the 16 types among the population, rather than use the frequencies provided for each type. For example, the fraction of the population that is INTJ comes to 0.029 (2.9 percent) when the values for I (0.507), N (0.267), T (0.402), and J (0.541) are multiplied. But the detailed table has INTJs as 2.1 percent of the population. In sum, there are discrepancies between the computed and given values of the 16 types in the population. The most striking discrepancy is for the INFJ type. When estimated from the frequencies of the four preferences, INFJs are 4.4 percent of the population; the table of values for all 16 types gives the percentage of INFJs as 1.5 percent.
Using the distribution given for the 16 types leads to somewhat different results: