I read a lot of climate papers so you don’t have to. Most of them are about the physical world, and you need some background to follow them. Every so often one comes along that requires no background at all, because it is about feelings, and we all have those. This is another one of those.
The paper is “Do climate emotions matter? Investigating their role in pro-environmental behavior,” published this month in the Journal of Environmental Psychology by Paula Blumenschein and colleagues at two German universities. They surveyed 966 Germans about their climate feelings, ran the answers through a regression, and concluded that climate feelings are important and should get more attention. I am going to walk you through it, because the walk is the fun part.
Let me start with the first sentence, because the first sentence does a lot of work:
The climate crisis is progressing at an alarming rate with serious implications for planetary and mental health.
That is the opening line. Not a hypothesis, not something the paper sets out to show. It is the floor the rest of the building sits on. The crisis is assumed, the alarming rate is assumed, and everything that follows is about how people feel inside that assumption. I want to flag this early, because if you do not share the premise that we are living through a planetary emergency, the study has already decided what that makes you. We will get to what that makes you. It is one of my favorite parts.
Here is the setup. The researchers used something called the Inventory of Climate Emotions, which measures eight separate climate feelings, each with four survey questions answered on a scale from strongly disagree to strongly agree. The eight feelings are climate anger, climate anxiety, climate guilt, climate sorrow, climate enthusiasm, climate powerlessness, climate contempt, and my personal favorite, climate isolation.
Climate isolation is defined as loneliness from other people not caring about climate change as much as you do. The survey measures it with statements like “I feel lonely because most of the people around me don’t care about climate change as much as I do.” This is now a measurable emotion with its own subscale. And in this study it turned out to be the single strongest predictor of whether someone goes to a climate protest. The lonelier you feel about caring more than your friends, the more likely you are to march. I could have told you that for free, but now it has a number, and the number is 0.30.
Now to my favorite part, which is climate contempt.
Contempt is the emotion the researchers built to capture climate skeptics. They define it as disregard for the issue of climate change. The actual survey question they use to measure it, and I am not making this up, is this:
It annoys me to watch people succumb to climate hysteria.
Sit with that for a second. Somebody who thinks a fair amount of climate alarm is overcooked, which describes most of the people I know and probably most of the people reading this, would agree with that sentence. The researchers have taken that view, which is a position about the world, and reclassified it as a negative emotion called contempt. You are no longer a person with an opinion. You are a person with a feeling, and the feeling has been entered into a statistics program.
And then a wonderful thing happened. When they ran the numbers, climate contempt came out predicting more climate activism, not less. The skeptics, in their model, were the ones going to the protests. This is the exact opposite of what the whole category was designed to show.
So what do you do when your denial-detector says the deniers are the activists? You call it a suppression effect. Here is their actual explanation:
This pattern should not be interpreted as evidence of a direct positive effect regarding those two outcomes, but emerges only when considered alongside other predictors and reflects suppression.
Suppression is a real thing in statistics. It is also what your results look like when you stuff eight nearly identical variables into the same model and the math starts arguing with itself. And these variables are nearly identical. The paper reports that climate anxiety and climate sorrow are correlated at 0.78. Anxiety and anger at 0.74. Anxiety and guilt at 0.70. These are not eight different emotions. This is one worried person being asked the same question eight slightly different ways, and the regression is now trying to tell the eight versions apart.
To their credit, and I mean that, the authors eventually admit it. They conclude that the contempt subscale “lacks construct validity,” which is the polite way of saying it does not measure what it was supposed to measure. They say the same about the powerlessness subscale, which also produced a backwards result. So two of the eight feelings are broken by their own account, and the other six are so tangled together the model is, in their own word, suppressing itself.
Now here is the part that I think actually matters, underneath the comedy.
The study has one outcome that is not a feeling. Most of what they measured was self-reported: how much activism do you do, how much do you nudge your friends. But for diet they used a carbon calculator on meat and egg consumption, which is at least an attempt to measure a real thing in the real world. And that is exactly where the whole apparatus falls apart. The feelings explained about 54 percent of the variation in self-reported activism, the squishiest measure, and about 17 percent of the variation in actual dietary emissions, the realest one. The closer they got to something physical, the less the feelings predicted it.
And climate guilt, which they expected would go with a cleaner diet, went the other way. The people who felt the guiltiest about their diet had the worst diets. Either guilt is a remarkable thing, or guilt follows your behavior instead of fixing it. The authors notice this, raise their eyebrows, and move along.
Which is the whole study, really. People who say they care a lot about climate change also say they do a lot about climate change. That is not a discovery. That is the same person answering two versions of the same question and the researchers drawing a line between the answers. The authors even admit, way down in the limitations, that the data cannot tell you which way the arrow points, whether the feelings drive the actions or the actions drive the feelings or both come from something else entirely. That is the correct thing to admit. It also quietly empties out the rest of the paper.
The conclusion, after all of this, is that climate groups should hold more sessions where people can share their feelings of climate isolation and climate enthusiasm. They reached that conclusion by surveying people in the climate movement and discovering the climate movement has feelings.
I do not doubt that people have real and powerful emotions about the environment. That is a genuine subject. But you cannot prove a crisis by assuming it in line one, you cannot turn one anxious mood into eight separate emotions by giving each its own subscale, and you certainly cannot take “it annoys me to watch people succumb to climate hysteria,” file it under contempt, and act surprised when the people who agree with it do not behave like the villains you cast them as.
The 966 Germans were asked how they felt. They told them. The feelings went into the model, the model made numbers, the numbers went into the journal, and everyone got paid, the survey panel included. And that, as far as I can tell, is the rest of the story.
