Income Inequality Itself Doesn’t Make People Unhappy

There’s an idea floating around the world of psychology referred to as Social Comparison Theory. The basic idea is that people want to know how well they measure up to others in the world and so will compare themselves to others. While there’s obviously more to it than that (including some silly suggestions along the lines of people comparing themselves to others to feel better, rather than to do something adaptive with that information), the principle has been harnessed by researchers examining inequality. Specifically, it has been proposed that inequality itself makes people sad. According to the status-anxiety hypothesis, when it comes to things like money and social status, people make a lot of upwards comparisons between themselves and those doing better. Seeing that other people in the world are doing better than them, they become upset, and this is supposed to be why inequality is bad. I think that’s the idea, anyway. Feel free to add in any additional, more-refined versions of the hypothesis if you’re sitting on them.

“People are richer than me,” now warrants a Xanax prescription

As it turns out, that idea seems to be a little less than true. Before getting to that, though, I wanted to make a few general points (or warnings) about the research I’ve encountered on inequality aversion; the idea that people dislike inequality itself and seek to reduce it when possible (especially the kind that leaves them with less than others). The important point to make about research on inequality is that, if you are looking to get a solid measure of the effects of inequality itself, you need to get everything that is not inequality out of your measures. That’s a basic point for any research, really. 

For instance, when examining research on inequality in the past, I kept noticing that the papers on the topic almost always contained confounding details which impeded the ability of the authors to make the interpretations of the data they were interested in making. Several papers looked at the effects of inequality on punishment in taking games. The basic set up here is that you and I would start with some amount of money. I then take some of that money from you for myself. Because I have taken from you, we either end up with you being better off, me being better off, or both of us being equal. After I take from you, you would be given the option to punish me for my behavior and, as it turns out, people preferentially punish when the taker ends up with more money than them. So if I took money from you, you’d be more likely to punish me if I ended up better off, relative to cases where we were equal or you were still better off. (This happens in research settings with experimental demand characteristics, anyway. In a more naturalistic setting when someone mugs you, I can’t imagine many people’s first thoughts are, “He probably needs the money more than me, so this is acceptable.”)

While such research can tell us about the effects of inequality to some extent, it cannot tell us about the effects of inequality that are distinct from takingTo put that in concrete example, my subsequent research (Marczyk, 2017) used that same taking game to replicate the results while adding two other inequality-generating conditions: one in which I could increase my payment with no impact on you, and another where I could decrease your payment at no benefit to myself. In those two conditions, I found that inequality didn’t appear to have any appreciable impact on subsequent punishment: if I wasn’t harming you, then you wouldn’t punish me even if I generated inequality; if I was harming you, you would punish me even if I was worse off. This new piece of information tells us something very important: namely, that people do not consistently want to achieve equality. When we have been harmed, we usually want to punish, even if punishing generates more inequality than originally existed. (That said, there are still demand characteristics in my work worth addressing. Specifically, I’d bet any effects of inequality would be reduced even further when the money the participants get is earned, rather than randomly distributed by an experimenter)

In terms of the research I want to talk about today, this is relevant because this new – and incredibly large – analysis sought to examine the effects of income inequality on happiness as distinct from the overall economic development of a country (Kelley & Evans, 2017). Apparently lots of previous work had been looking at the relationship between inequality within nations and their happiness without controlling for other important variables. The research question of this new work was, effectively, all else being equal, does inequality itself tend to make people unhappy? The simple example of this question they put forth was to imagine twins: John and James. John lives in a country with relatively low income-inequality and makes $20,000 a year. James lives in a country with relatively high income-inequality and makes $20,000 a year. Will John or James be happier? They sought to examine, on the national level, this connection between inequality and life satisfaction/happiness. 

“At least we’re all poor together”

In order to get that all else to be equal, there are a number of things you need to control for that might be expected to affect life satisfaction. The first of these is GDP per capita; how much a nation tends to produce per person. This is important because it might mean a lot less for your happiness that everyone is equal if that equality means everyone lives in extreme poverty. If that happens to the be the case, then increasing industrialization of a nation can actually increase opportunities for economic advancement while also increasing inequality (as the rewards of such a process aren’t shared equally among the population, at least initially. After a time, a greater percentage of the population will begin to share in those rewards and the relationship between inequality and economic development decreases).  

The other factors you need to control for are individual ones. Just because a society might be affluent, that does not mean that the person answering your survey happens to be. This means controlling for personal income as well, as making more money tends to make for happier people. The authors also controlled for known correlates of happiness including sex, age, marriage status, education, and religious attendance. It’s only once all these factors have been controlled for that you can begin to consider the effect of national inequality (as measured by the Gini coefficient) on life satisfaction ratings. That’s not to say these are all the relevant controls, but they’re a pretty good start. 

Enacting these controls is exactly what the researchers did, pooling data from 169 surveys in 68 societies, representing over 200,000 individuals. If there’s a connection between inequality and life satisfaction to be found, it should be evident here. Countries were categorized as a member of either developing nations (those below 30% of the US per-captial GDP) or advanced ones (those above the 30% mark), and the same analyses was run on each. The general findings of the research are easy to summarize: in developing nations, inequality was predictive of an increase in societal happiness (about 8 points on a 1-100 scale); among the advanced nations, there was no relationship between inequality and happiness. This largely appeared to be the case because, as previously outlined, the onset of development in poorer countries generated initial periods of greater inequality. As development advances, however, this relationship disappears.

A separate analysis was also run on families in the bottom 10% of a nation in terms of income, compared with the families in the top 10% since much of the focus on inequality has discussed the divide between the poor and the rich. As expected, rich people tended to be happier than poor ones, but the presence of inequality was, as before, a boon for happiness and life satisfaction in both groups. It was not that inequality made the poor feel bad while the rich felt good. Whatever the reason for this, it does not seem like poor people were looking up at rich people and feeling like their life was terrible because others had more.

“Some day, all this could be yours…”

All this is not to say that inequality itself is going to make people happy as much as the things that inequality represents can. Inequality can signal the onset of industrialization and development, or it can signal there is hope of improving one’s lot in life through hard work. These are positives for life satisfaction. Inequality might also represent that the warlord in the next town over is very good at stealing resources. This would be bad. However, whatever the reason for these correlations, it does not seem to be the case that inequality per se is what makes people unhappy with life (though living in nations with high GDP and earning good salaries seem to put a smile on some faces).

I like this interpretation of the data, unsurprisingly, because it happens to fit well with my own. In my experiments, people didn’t seem to be punishing inequality itself; they were punishing particular types of behaviors – like the stealing or destruction of resources – that just so happened to generate inequality at times. In other words, people are responding primarily to the means through which inequality arises, rather than the inequality itself. This appears to be the case in the present paper as well. Most telling of this interpretation, I feel, is a point mentioned within the paper without much discussion (as its the topic of a separate one): the national data was collected from non-communist nations. Things are a little different in the communist countries. For those cohorts who lived their formative years in communist nations, inequality appears to have a negative relationship with happiness, though that dissipates in new, post-communist generations. From that finding, it seems plausible to speculate that communists might have different ideas about the means through which inequality arises (mostly negative) which they push rather aggressively, relative to non-communists. That said, those attitudes do not seem to persist without consistent training.

Reference: Kelley, J. & Evans, M. (2017). Societal inequality and individual subjective well-being: Results from 68 societies and over 200,000 individuals, 1981-2008. Social Science Research, 62, 1-23.

Marczyk, J. (2017). Human punishment is not primarily motivated by inequality. PLOS One, https://doi.org/10.1371/journal.pone.0171298

Semen Quality And The Menstrual Cycle

One lesson I always try to drive home in any psychology course I teach is that biology (and, by extension, psychology) is itself costly. The usual estimate on offer is that our brains consume about 20% of our daily caloric expenditure, despite making up a small portion of our bodily mass. That’s only the cost of running the brain, mind you; growing and developing it adds further metabolic costs into the mix. When you consider the extent of those costs over a lifetime, it becomes clear that – ideally – our psychology should only be expected to exist in an active state to the extent it offers adaptive benefits that tend to outweigh them. Importantly, we should also expect that cost/benefit analysis to be dynamic over time. If a component of our biology/psychology is useful during one point in our lives but not at another, we might predict that it would switch on or off accordingly. This line of thought could help explain why humans are prolific language learners early in life but struggle to learn a second language in their teens and beyond; a language-learning mechanism active during development it would be useful up to a certain age for learning a native tongue, but later becomes inactive when its services are no longer liable to required, so to speak (which they often wouldn’t be in an ancestral environment in which people didn’t travel far enough to encounter speakers of other languages).

“Good luck. Now get to walking!”

The two key points to take away from this idea, then, are (a) that biological systems tend to be costly and, because of that, (b) the amount of physiological investment in any one system should be doled out only to the extent it is likely to deliver adaptive benefits. With those two points as our theoretical framework, we can explain a lot about behavior in many different contexts. Consider mating as a for instance. Mating effort intended to attract and/or retain a partner is costly to engage in (in terms of time, resource invest, risk, and opportunity costs), so people should only be expected to put effort into the endeavor to the extent they view it as likely to produce benefits. As such, if you happen to be a hard “5″ on the mating market, it’s not worth your time pursuing a mate that’s a “9″ because you’re probably wasting your effort; similarly, you don’t want to pursue a “3″ if you can avoid it, because there are better options you might be able to achieve if you invest your efforts elsewhere.

Speaking of mating effort, this brings us to the research I wanted to discuss today. Sticking to mammals just for the sake of discussion, males of most species endure less obligate parenting costs than females. What this means is that if a copulation between a male and female results in conception, the female bears the brunt of the biological costs of reproduction. Many males will only provide some of the gametes required for reproduction, while the females must provide the egg, gestate the fetus, birth it, and nurse/care for it for some time. Because the required female investment is substantially larger, females tend to be more selective about which males they’re willing to mate with. That said, even though the male’s typical investment is far lower than the female’s, it’s still a metabolically-costly investment: the males need to generate the sperm and seminal fluid required for conception. Testicles need to be grown, resources need to be invested into sperm/semen production, and that fluid needs to be rationed out on a per-ejaculation basis (a drop may be too little, while a cup may be too much). Put simply, males cannot afford to just produce gallons of semen for fun; it should only be produced to the extent that the benefits outweigh the costs.

For this reason, you tend to see that male testicle size varies between species, contingent on the degree of sperm competition typically encountered. For those not familiar, sperm competition refers to the probability that a female will have sperm from more than one male in her reproductive tract at a time when she might conceive. In a concrete sense, this translates into a fertile female mating with two or more males during her fertile window. This creates a context that favors the evolution of greater male investment into sperm production mechanisms, as the more of your sperm are in the fertilization race, the greater your probability of beating the competition and reproducing. When sperm competition is rare (or absent), however, males need not invest as many resources into mechanisms for producing testes and they are, accordingly, smaller.

Find the sperm competition

This logic can be extended to matters other than sperm competition. Specifically, it can be applied to cases where a male is (metaphorically) deciding how much to invest into any given ejaculate, even if he’s the female’s only sexual partner. After all, if the female you’re mating with is unlikely to get pregnant at the time, whatever resources are being invested into an ejaculate are correspondingly more likely to represent wasted effort; a case where the male would be better off investing those resources to things other than his loins. What this means is that – in addition to between-species differences of average investment in sperm/semen production – there might also exist within-individual differences in the amount of resources devoted to a given ejaculate, contingent on the context. This idea falls under the lovely-sounding name, the theory of ejaculate economics. Put into a sentence, it is metabolically costly to “buy” ejaculates, so males shouldn’t be expected to invest in them irrespective of their adaptive value.

A prediction derived from this idea, then, is that males might invest more in semen quality when the opportunity to mate with a fertile female is presented, relative to when that same female is not as likely to conceive. This very prediction happens to have been recently examined by Jeannerat et al (2017). Their sample for this research consisted of 16 adult male horses and two adult females, each of which had been living in a single-sex barn. Over the course of seven weeks, the females were brought into a new building (one at a time) and the males were brought in to ostensibly mate with them (also one at a time). The males would be exposed to the female’s feces on the ground for 15 seconds (to potentially help them detect pheromones, we are told), after which the males and females were held about 2 meters from each other for 30 seconds. Finally, the males were led to a dummy they could mount (which had also been scented with the feces). The semen sample from that mount was then collected from the dummy and the dummy refreshed for the next male.

This experiment was repeated several times, such that each stallion eventually provided semen after exposure to each mare two or three times. The crucial manipulation, however, involved the mares: each male had provided a semen sample for each mare once when she was ovulating (estrous) and two to three times when she was not (dioestrous). These samples were then compared against each other, yielding a within-subjects analysis of semen quality.

The result suggested that the stallions could – to some degree – accurately detect the female’s ovulatory status: when exposed to estrous mares, the stallions were somewhat quicker to achieve erections, mount the dummy, and to ejaculate, demonstrating a consistent pattern of arousal. When the semen samples themselves were examined, another interesting set of patterns emerged: relative to dioestrous mares, when the stallions were exposed to estrous mares they left behind larger volumes of semen (43.6 mL vs 46.8 mL) and more motile sperm (a greater percentage of active, moving sperm; about 59 vs 66%). Moreover, after 48 hours, the sperm samples obtained from the stallions exposed to estrous mares showed less of a decline of viability (66% to 65%) relative to those obtained from dioestrous exposure (64% to 61%). The estrous sperm also showed reduced membrane degradation, relative to the dioestrous samples. By contrast, sperm count and velocity did not significantly differ between conditions.

“So what it was with a plastic collection pouch? I still had sex”

While these differences appear slight in the absolute sense, they are nevertheless fascinating as they suggest males were capable of (rather quickly) manipulating the quality of the ejaculate they provided from intercourse, depending on the fertility status of their mate. Again, this was a within-subjects design, meaning the males are being compared against themselves to help control for individual differences. The same male seemed to invest somewhat less in an ejaculate when the corresponding probability of successful fertilization was low.

Though there are many other questions to think about (such as whether males might also make long-term adjustments to semen characteristics depending on context, or what the presence of other males might do, to name a few), one that no doubt pops into the minds of people reading this is whether other species – namely, humans – do something similar. While it is certainly possible, from the present results we clearly cannot say; we’re not horses. An important point to note is that this ability to adjust semen properties depends (in part) on the male’s ability to accurately detect female fertility status. To the extent human males have access to reliable cues regarding fertility status (beyond obvious ones, like pregnancy or menstruation), it seems at least plausible that this might hold true for us as well. Certainly an interesting matter worth examining further.   

References: Jeannerat, E., Janett, F., Sieme, H., Wedekind, C., & Burger, D. (2017). Quality of seminal fluids varies with type of stimulus at ejaculation. Scientific Reports. 7, DOI: 10.1038/srep44339