I started Do the Math in 2011 as a way to reach a larger audience than a handful of students every year or two in an energy course at UCSD. I had (and still have) deep concerns about the assumptions we make as a society based on our fossil fuel trajectory over the last century or so. Trying to steer policy from the top seemed a losing proposition: feckless politicians hew to their constituents’ desires via a mechanism we call democracy, so why not try to get people on board directly?
I never imagined creating a blog that would get millions of pageviews, although this by itself falls well short of having an impact on a grand scale. But I figured I owed it to myself to reach as many as I might. What I have found is that a select few seem to share my concerns. And some vocal contributors to comments strongly disagree that we need to worry (why then make the seemingly wasted effort to respond to—in their eyes—doomsaying kooks if in fact we need not be concerned?). But most people simply don’t care enough to tune in. They’ve learned to ignore prognostications of any flavor, perhaps. Lately, even fewer people are entertaining ideas of resource limits owing to increased global oil production (led almost entirely by U.S. shale oil) and a recovering economy.
But I think there is something more fundamental going on here. I think we’re dealing with personality traits cooked into human nature. Are we capable of mitigating a far-off potential calamity via proactive efforts decades ahead of a putative crisis? In this post, I’ll use some survey data suggesting that we may be in trouble.
Evidence of a Problem
About a year ago, a friend shared with me a graphic from an informal survey on the Peak Prosperity website. This is Chris Martenson’s site, which hosts a “Crash Course” consisting of 4.6 hours of quality video content describing why we should worry that tomorrow may not be bigger than today, and why the growth phase may be just that. As a related aside, I once did a podcast interview for Chris Martenson.
The Peak Prosperity site visitors probably have a lot of overlap with Do the Math readers: the fundamental concern is the same. These are people who are—by and large—not content with extrapolation of the here-and-now into tomorrow. We think there will be fundamental changes in how the full Earth operates compared to our frontier days of resource exploitation in an empty Earth. In many cases, there are compelling calculations to motivate concern. Rather than trying to predict a dire future, my goal in “Did the Math” was to build a plausible case for things going off the rails in the desperate hope that recognition of this possibility would spur action now to steer clear of this potential pitfall (thereby making me wrong, in a happy way). It’s trying to expose a blind spot—a sleeping dragon.
But that blind spot may be stamped into human nature. So what about this survey?
The survey asked people to indicate their personality type as classified by the Myers-Briggs mechanism. Keep in mind that these are people visiting the Peak Prosperity forum. We’ll call these people “receptive to the cautionary message.” Or at least engaged in the issue—be they supporters or detractors.
The Survey Result
I’ll define the types in a bit, but first let’s look at the data from Peak Prosperity.
We see a huge spike for the INTJ type. Want to guess what type I am?
The result was pretty stunning. Of the 114 responses, site visitors were dominated by INTJ types (43 in number, or 38%), even though this group constitutes about 2–3% of the population. The website appears to be highly selective. It’s as if you called a meeting in San Diego to discuss drill bits and almost half the attendees were red-heads. If accurate, the implication is that less than 8% of the entire human population is likely receptive to the cautionary message on Peak Prosperity (and by extension, Do the Math—the numbers from which suggest an even smaller number). That’s a small fraction of the population, and likely well short of a “critical mass” for preventive action. So we may be committed to crisis.
Motivated by this stunning result, I went to Do the Math readers to get another survey of the same nature—but this time with a much greater degree of control and knowledge about the methods. I got almost 1000 unique responses, and the result is strikingly similar—if not even more skewed. I’ll defer showing it until a bit later. We will first take a detour to explain the Myers-Briggs indicator, and then tease out some interesting results from the new data.
Myers Briggs Types
I can’t add to what’s been said about MB types, and recommend the Wikipedea page for background. I will, however, offer a brief elucidation so that you don’t have to run off elsewhere to at least master the lingo.
The MB indicator has four fields/characteristics that each contain one of two possible letters/traits. Thus there are 16 possible combinations/types, which would result in 6% per type if uniformly distributed (not the case, of course). Every person has some of both polar characteristics, and some individuals will be kind-of in the middle (not convincingly to one side or the other).
The first field is either an I or E depending on whether a person tends to feel depleted or energized by social interaction.
The second field is S or N. Does a person rely mostly on direct sensory input for their information, or does it come more from a synthesized, abstract, intuitive read of a larger set of inputs? Think of this one as concrete vs. abstract tendencies. Someone who holds a snowball in front of Congress to argue against climate change is a hard-over S-type. The S/N dichotomy is perhaps the most important attribute when it comes to whether an individual would entertain Do the Math views.
The T/F dichotomy is pretty self-explanatory. Does a person place more stock in cold logic and deduction rather than in consideration of other peoples’ needs and feelings? Spock or Kirk?
The J/P designation is the hardest for me to describe. Judging does not mean judgmental. Technically, J-types rely more heavily on their T/F facility and P-types rely more on S/N when it comes to processing information from the world. As an emergent property, combine T with J (like me) and you have a punctual planner on your hands.
Put these four traits together and you make people. There are loads of pages and books describing attributes of the MB types. The official Myers-Briggs test costs money to take. But there are plenty of knock-offs you can find on the web (like here—be aware that scoring 56% in a category means 56% of the way from neutral to hard-over; or here; or here). If you want to see a pretty accurate description of the INTJ type, for instance, see this link (and edit last four letters of URL to see other types).
It can be fascinating to read about the various types. When I read about INTJ, I find myself saying: “Yup, they’ve got me—how did they even know that about me?” It’s not like horoscopes with lots of generalities such that anyone might identify with significant portions of any description. When I read other Myers-Briggs type descriptions, I find far less overlap. And unlike astrology, which ridiculously ties personality traits to time of birth, the MB type is based on your personal, presumably honest responses to a battery of questions: the result is self-directed to be a personally relevant/accurate reflection of what makes you tick.
And just pause to consider: do you think a poll of Peak Prosperity or Do the Math readers would show a correlation or strong preference for astrological sign or birth month? When any type classification scheme produces such dramatic results as we see above, there must be something “real” to the scheme—even if you do not understand what it is. The fact that so many scientists are INTJs is no accident. Personality type matters, and often manifests very early in life.
The Wikipedia page has some stats on the prevalence of the 16 types in the U.S. population, as well as some temperament labels. Other sites have their own labels; the table below provides some synthesis.
|MB Type||Prevalence Range||Adopted Prevalence||labels|
|ISTJ||11–15||13, 12||inspector, logistician|
|ISTP||4–6||5, 5||craftsman, virtuoso|
|ISFJ||9–15||12, 13||protector, defender|
|ISFP||5–9||7, 9||adventurer, composer|
|INTJ||2–4||3, 2||mastermind, architect|
|INTP||3–5||4, 3||architect, logician|
|INFJ||1–3||2, 2||counselor, advocate|
|INFP||4–5||4.5, 4||healer, mediator|
|ESTJ||8–12||10, 9||supervisor, executive|
|ESTP||4–5||4.5, 4||promotor, entrepreneur, dynamo|
|ESFJ||9–13||11, 12||provider, consul|
|ESFP||4–9||6.5, 9||performer, entertainer|
|ENTJ||2–5||3.5, 2||commander, field marshal|
|ENTP||2–5||3.5, 3||inventor, debater, visionary|
|ENFJ||2–25||3.5, 3||teacher, protagonist|
|ENFP||6–8||7, 8||champion, campaigner|
For full disclosure, I tweaked the high end of the ISTJ and ISFJ ranges by 1% each so that the sum of average values would come out to 100%. In the “adopted prevalence” column, I put two numbers: the first is the average of the range shown in the second column, and the other comes from the Truity site (which also has gender breakdown).
Do the Math Results
So I put up a survey invitation on Do the Math to see if the Peak Prosperity result held up. I wanted to make the survey super-simple (one question with 16 drop-down choices), but one day and hundreds of responses later, I came to wish that I had asked a couple more questions (Did you already know your MB type?; Do you subscribe to the cautionary message?). So I made another appeal and had 230 do-overs plus about 500 new respondents for a total of 958 (725 in total responded to the second poll with additional questions).
There are no statistical alarm bells in comparing the two sets of responses, so I present the combined data set here.
Here is the result, plotted on top of the prevalence data (Wikipedia in red, Truity in blue, overlap appears purple).
Pretty much the same story as in the original Peak Prosperity poll. The main difference is that DtM lacks as many INFJ as PP. Maybe DtM’s analytic, quantitative bent selects T over F people somewhat more. But in any case, note that the DtM personality distribution and the general population look nothing alike! Look at all that red hair in the audience! Who knew drill bits were so exclusively interesting to that crowd?
Okay, now that we have some of the basics about types and prevalence under our collective belt, and have seen the Do the Math data, let’s assess what the results mean.
INTJ represented 44% of the survey respondents, while only constituting 2–3% of the population. That is, very simply (and to repeat), what is so notable about this result. I still can’t get over it.
Playing off of this extreme skew, let’s pretend that 100% of INTJ types would listen to the Do the Math message and take it seriously (persuadable). This group has already shown a predilection: let’s take it to the extreme. Admittedly, this may be opening the barn door too wide, but we’ll see that even this generous assumption results in a very small barn. We can then calculate the interest level in other groups.
The following graph helps illustrate the approach I am about to take.
In this representation, I have plotted the Wikipedia-obtained prevalence in pink. These add to 100%. I overplot the DtM survey response in such a way that the blue line does not overtop the population in any bin. In other words, we can’t gather more of any type than exist in society. Not surprisingly, the wildly over-represented INTJ type sets the limit on saturation. Setting the two distributions to be equal for INTJ is equivalent to claiming that 100% of INTJs are potential adherents to Do the Math (that they would all be natural followers/listeners to the cautionary message). Again, this is likely an overstatement.
But here’s where it gets interesting, in two ways. First, add all the DtM people up on this scale and now we only get 6.8% of the population. That’s a bit less than the 8% I get when I do the same thing using the Peak Prosperity results. This number becomes 4.5% if I use the Truity prevalence numbers. So it’s in the 5–6% ballpark.
The second interesting thing is that we can determine the interest level of other types (scaled to 100% for INTJ: reduce all numbers if this is not accurate). Looking at the plot above, we can see that INTP is about 40% represented, INFJ is maybe 25% represented, and down from there. So here is a plot of deduced interest level for the two sets of prevalence data (same color scheme as before).
Where did ESXX go? And ISFX?
Remember, this is all normalized to the assumption that 100% of INTJs could get on board. To the extent this is not true, the number in the total population goes down as well.
The result is noteworthy. Even if off by a factor of two due to some systematic problem (explored below), the upshot is that we probably don’t have a high enough fraction of people with the disposition to take the cautionary message seriously, in advance of evident crisis. If 5% is too low to be a critical mass (as I suspect it is), then this could spell our doom: human nature is not up to the challenge.
Cassandra—while being right in the end—is invoked as a label of unjustified alarmism. “Crying wolf” is used to indicate empty warnings, even though the wolf in the story did come in the end. Isn’t it telling, the lesson we take away from these stories? Rather than conclude that we should always examine warnings (the consequences are bad), we create labels to admonish the purveyors of alarm.
If the cautionary message is way off base, then perhaps it’s just as well that the adherents never reach critical mass—though I’m not sure what damage would result from an aggressive campaign to reduce depletion of fossil fuels and other resources, other than economic slowdown/reversal. Of course if you buy into the utopian dream this is criminal enough—depriving countless individuals of comfort and easy-living.
You Call this Science?
Well, no. I’m not calling what I’m doing here “science.” It’s analysis, based on a couple of selective sets of survey data. It’s interesting. It has a significant chance of being right, guided by some intuition I’ll discuss in a bit. At the very least it seems to be something we should pay attention to, and if nothing else try to do a more controlled scientific study. I think it’s important. If humans as a whole are wired not to plan far ahead for contingencies of a sort never before seen, then we should know this about ourselves and acknowledge a fundamental vulnerability (other suitable words here are: shortcoming; liability; blind-spot).
Could something be wrong with the data?
My first exposure to this amazing skew was via the Peak Prosperity forum survey for which I had no personal knowledge of the source and methods employed: was there a link to take a MB type test in case it wasn’t already known? Were we just seeing a reflection of who already knew their types, and the INTX folks really go in for this sort of thing? Do INTX people love taking surveys (I don’t, particularly)? The result was so striking I also contemplated the possibility that it was fabricated to make a point. I didn’t really think so, but could not truly say.
This is why I took on the survey myself. In doing so, I know the audience (Do the Math readers who like the site well enough that either they set up notifications for new posts, or dedicated individuals who check in periodically for increasingly rare posts). I also know that I provided a link to a test in case people didn’t already know their type.
I saw the same pattern emerge (though with markedly less INFJ), but still had doubts. Was I still just getting people who already knew their type? So a couple days later I put up a new survey to supersede the original, adding two questions: Did you already know your type before taking the DtM poll?; and essentially: Do you subscribe to the DtM cautionary message?
I had 463 responses to the first type-only survey. Via the second survey (725 responses), I found that about 35% of people already knew their MB personality type, but no evidence that this was responsible for overstocking INTJ responses. The only statistically significant deviations from the typical level were that INTPs tend to know who they are, and INFJs don’t. None of the other deviations (shown below) have the numbers to be differentiable from random.
The other thing I found was that 76% of respondents agreed with the statement: “I worry that we risk systemic failure if we don’t acknowledge limits to resources and growth.” A small 6% opted for the statement: “Human innovation and economic markets will avert any collapse scenario.” 18% said that neither statement captured them well. I did not see compelling evidence that the optimist camp tends toward certain types (although ISTJ almost stands out here). Note, however, that most types are not well represented in this poll: the non-INTX types are oddball outliers among their fellow types as DtM visitors, so it is risky to draw conclusions about the attitudes of non-INTX folks with respect to the cautionary message. The only statistically significant anomalies are that INTPs are less likely to be pinned down (26% rejected both statements), and 17% of ISTJs liked the optimist answer.
Various selection effects could be operating here to skew the PP and DtM results so markedly. Maybe INTJs, being analytical creatures, are more likely to know about Myers-Briggs and know their personality type. I did (but the result described above does not bear this out). Maybe INTJs were more likely to see value in responding to the survey. Maybe INTJs are more likely to participate in online forums of any type. Maybe the way word spreads about online forums involves consort selection effects like co-workers, where correlations exist (I, for instance have no shortage of INTJ types around me at work). Maybe all these things are playing a role at once.
I’m sure some of these effects are distorting the numbers. But is it a whole factor of two? Three, even? Even at this level, we still have < 20% of people on board with the message (again under the unlikely assumption all INTJs would be). My sense is that this still fails to meet critical mass for large-scale action in a non-crisis democracy, any way we slice it.
Being an INTJ, I rely more on intuition and abstraction than on what I sense around me in the here-and-now. Scientists must do so in order to develop theoretical frameworks. I have a healthy appreciation for the concrete (experimentalist in me), but do still lean on abstraction. I have noticed that I tend to generalize situations looking for the common lesson. I like to synthesize. It’s the distilled product that sticks with me in long term memory. I am prone to make a statement based on a wide variety of inputs over the years, but when challenged to cite specific examples, I struggle to recreate them: I’ve already chucked them out in favor of the overarching principle.
The sensing/intuiting split is the most crucial personality component in the present context, although thinking/feeling is a very important dimension as well. In order to project current practices into the future and recognize a peril that does not look very much like the world generations have known requires abstraction (and thinking), rather than sensing the immediate world. So N types are most likely to heed the cautionary message. Look again at the “interest” figure and this bears out extremely well. I don’t see that as an accident.
According to the prevalence data, S types constitute 69–73% of people. That’s a lot! And look at how non-represented they are on the “interest” plot. We’re missing most of the people. This is part of why it is also so difficult to get people on board with global warming messages. Thinkers are more common, though still in the minority, at 40–47%. So it’s the S/N characteristic that’s the bottleneck.
Even if the survey was deeply flawed, such that in reality 20% of S-types were responsive to the cautionary message (as opposed to 1% reported) and as many as 50% of the N-Types, on average, were receptive (twice as large as shown), we would still have less than 30% representation. And this is a pretty severe systematic distortion of the poll. If 100% of N-types are receptive and 0% of S-types, we’re also at 30%. It is hard to imagine selection effects being so completely slanted toward INTJ that these distortions would be allowable. And if this were the case, we would be forced to put even more stock in the usefulness/meaningfulness of personality types.
In summary, if you base your world view on concrete, “provable” facts sitting in plain sight, then you’ll have little truck with future extrapolations and musings. The more radically the projection departs from the “normal,” the less likely it can be taken seriously. These are good, solid traits for humans to have—and we can imagine their merit in an evolutionary sense. Things usually are as they are and have been throughout living memory. Except when they’re not.
Bring on the Extraverts!
One other dimension worth poking: what if, for instance, an ENTJ is every bit as likely to be persuaded to heed the cautionary message as an INTJ, except they spend their time socializing (in person or virtually) rather than reclusively reading scary websites? What if they could be brought on board? Indeed, the interest level among ENXX types somewhat mirrors the descending staircase seen in the INXX group in the Interest Level plot. What if the extraverts were boosted (by a factor of 13, it turns out) so that 100% of ENTJs would be receptive?
The result is in the extravert-tweaked figure below, using the (more generous) Wikipedia-obtained prevalence numbers.
Welcome aboard, party people! Now we are 11–18% of the population, depending on prevalence numbers—up from 4.5–6.8% when you were busy socializing. Hmm. Still not an impressive showing. Still sub-critical—even if this distortion/boost is justified.
As stated at the beginning, my impetus for all the Do the Math work was to lay out a rational, quantitative foundation for why we should not take future growth/wealth/happiness for granted. We could really blow this thing. Our best hope, as I saw it, was to get people to acknowledge and accept the threat and thereby endeavor to make it go away. As with any 12-step program, admitting that there is a problem is step one.
Failure to acknowledge what, to me, is a wholly plausible set of major concerns triggers a strong reaction on my part. How can we mitigate what we don’t acknowledge? Failure to acknowledge the risk serves only to solidify the likelihood of the risk, in my mind. Perversely, calling me wrong outright probably makes me more right. Saying I might be wrong, or even that I am probably wrong while admitting some chance of my being on target and acknowledging the enormity if so is just fine by me.
This personality analysis helps me understand the scope of the challenge. It mostly serves to reinforce my concern. It seems we have a built-in impediment to preventive mitigation for unprecedented crises. At some level, it just makes me feel resigned: no hope in politicians, now no hope in human nature.
But as a cerebral type, it gives me some satisfaction to have insight into how and why we may fail. If the world falls apart before I die, at least I’ll have some inkling as to what’s going on, and won’t be as psychologically shattered by the affair. But I’ll be one of a pitifully small number, I’m afraid.