Sunday 14 January 2018

Multi-disciplinary thinking; the gender wage gap; and amoral markets

I am a big fan of multi-disciplinary thinking. I think it leads to vastly superior judgement, and in the field of investing, superior judgement is the cornerstone of generating superior returns. Investing is a competitive pursuit that requires one to have superior insights to one's competitors in the market, and multi-disciplinary thinking - because it is so rare and difficult - can act as an important competitive advantage in this regard. It is also important in the field of policy analysis and in many other fields where complex judgement is required (but unfortunately, is too frequently lacking).

We live in a world of specialists. From the time we begin to attend university, we specialise. And most people then move on and work in highly specialist fields. Consequently, most people have a strong grasp of a narrow specialty, but often suffer from a profound degree of ignorance with respect to even the basic fundamentals of alternative fields - even within their broader discipline (e.g. an employment lawyer may know a lot about the intricacies of employment law, but next to nothing about tax law).

However, the reality of the world does not observe the bounds of traditional university departments or vocational specialties. All specialties are individual pieces of fabric that are woven together into a unifying whole that represents the reality in which we inhabit, and being familiar with only one discipline is a sure path to being unable to see the big picture or exercise sound judgment. There is a very old saying - to a hammer everything looks like a nail - and it is a profoundly common problem in modern day life.

How is it possible for one individual to understand and unify multiple disciplines? Don't we live in a world of exponentially increasing knowledge and data. Every minute, 300hrs of new content is uploaded to YouTube. It would be literally impossible for one individual to ever watch more than a tiny fraction of it. The answer is this: while the amount of data is mushrooming exponentially, the amount of underlying truth is finite and humanity's expansion in accumulated knowledge ought to be increasingly unifying. A large fraction of the growing availability of data is not adding new knowledge - there is more noise than signal being generated. The more you already know, the easier it is to integrate new knowledge, in much the same way as a jigsaw puzzle gets easier the more pieces you already have in place. Understanding the world broadly is like completing a jigsaw puzzle.

The core reason multi-disciplinary thinking works so well is that it allows an observer to simultaneously bring to bear several mental models to the analysis of a problem. The advantage is that these mental models can be then triangulated with one another to see whether they are all pointing towards an internally-consistent conclusions. If they are all tending towards the same conclusion, one can have a high degree of confidence that their conclusion is probably correct. However, perhaps more importantly, if they are not all pointing to the same conclusion, it can alert one to the reality that the problem is far more complex and nuanced than it might first appear. In this respect, multi-disciplinary thinking not only provides additional confidence in ones positive conclusions, but it also provides an important warning as to when one's perspective may well be wrong.

Practitioners that have only one set of mental models from one discipline will tend to see the world in black and white terms. If observations appear to conform with the dictates of the mental models at the core of their narrow discipline, they have no way to cross-check their conclusions against mental models deriving from other disciplines. This can often lead to a toxic combination of high confidence and poor judgment, borne of a blinkered, one-dimensional view of the world. And as Mark Twain once said, "It ain't what you don't know that gets you into trouble. It's what you know for sure that just ain't so".

Highly specialised, one-dimensional perspectives are the origin of ideology, in my view. Ideology comes from having only a narrow set of mental models to explain the world, and it leads to highly constrained and error-prone judgment. When someone with one set of narrow mental models and ideology clashes with someone with another set of narrow mental models and ideology, frequently the outcome is frustration, anger, disagreement, and mutual views that the other person is an idiot and don't know what they are talking about. And they are both right - both sides are indeed idiots.

Investing great and Warren Buffett sidekick Charlie Munger has long talked about the virtues of having various mental models. He believes himself to have roughly 100 spanning a large number of disciplines. He is absolutely on to something here in my view. Further support comes from the book Superforecasting: The art and science of prediction, which is one of the best books I have ever read. The authors Tetlock and Gardner note that one of the key attributes that all unusually-successful forecasters have in common is that they are broad, multi-disciplinary thinkers rather than narrow field experts. In light of the above, we ought not view that as surprising.

Applying the above to the topical issue of the gender wage gap

An interesting example of how multi-disciplinary thinking can lead to superior insights to narrow ideological perspectives is the long-standing and much-discussed issue of the gender pay gap. The forgoing discussion brings to bear three different perspectives on this issue, before attempting to unify them: the sociological perspective; the economics perspective; and the evolutionary psychology/biological perspective.

There is clear statistical evidence that on average, men are paid more than women, with the gap being approximately 20% (before controlling for various factors). That much is agreed upon. The question is, how should we interpret and explain this outcome, and what should we prescriptively do from a policy perspective, if anything?

#1 The sociological perspective

Observers from left-leaning social science departments (particularly sociology) have a certain framework and set of mental models about the way the world works that they use to analyse the problem. To their way of thinking, we live in a world full of systemic discrimination in which the powerful oppress the weak. We live in a world with a long history of patriarchal oppression of women (evidenced by women's historical lack of rights to vote; own property; or have the same access to educational and vocational opportunities as men), as well as oppression by whites in Western society of ethnic minorities (which finds an expression in the notion of 'white privilege'). Evidence for the latter dates back to the era of slavery, as well as the systemic discrimination African Americans continued to suffer in the years post abolition up until at least the 1960s civil rights movement.

It is argued that although much progress has been made in improving the rights of women and ethnic minorities over the past century, that this history continues to cast a long shadow, and that systemic and implicit bias against women and ethnic minorities continues to persist until this day.

To someone with this world view, the cause of the gender wage gap is obvious - the patriarchy is still alive and well, and men - and in particular white men - continue to discriminate unfairly against women. This finds expression in women being systematically provided with less opportunities for career advancement, and/or being underpaid vs. men for the same work. It is an open and shut case, and compelling evidence for the continuing societal discrimination against women.

The policy prescriptions stemming from this worldview are seemingly obvious: we should implement more activist policies to ensure female representation in high paying fields; try to actively promote more women into leadership roles; and enforce stricter standards for gender pay parity (as well as equal gender representation in key roles). Failure to achieve gender parity and/or close the pay gap will represent evidence of continuing discrimination against women, requiring both individual offenders to be called out and held to account, as well as more vigorous remedial policy action.

#2 The economics perspective

But here is where it starts to get interesting. Let's now take off our sociological caps for a moment and head over to the economics department, and don a new economics cap. Now we will be addressing the problem with a completely different set of mental models.

An economics approach to the gender wage gap would come at the problem from the perspective of efficient labour markets. If the labour market is efficient, in the aggregate, workers will be paid approximately in accord with their productivity and bargaining power, whereas if labour markets are inefficient, they would not. Economic theory predicts that labour markets will be quite efficient, because the combination of the profit motive and the significant economic rewards available to those able to successfully exploit labour market inefficiencies would be considerable.

A 20% underpricing of female labour would be an extraordinary labour market inefficiency. If that was indeed the case, what would an economics perspective predict would happen? Well, the first prediction would be that if a company were to set up in business and employ only women, that it would have a substantial labour cost advantage over its competitors - particularly in labour-intensive industries. These businesses would then be more profitable than competing businesses that employed more men, and so would therefore have higher profit margins and/or lower costs, and would therefore have more funds available to finance investment and/or capacity to reduce prices to capture market share. The economics perspective would predict that companies employing more women would be more profitable; would be growing faster; and would be thriving in particular in labour-intensive fields.

Do we see any evidence of that in the capitalist free market, where open competition exists, and millions of businesses jostle with each other daily for any small or fleeting advantage? Do we see any evidence for it in a world where women are free to start their own businesses and hire other women? Do we see that happening anywhere in the world, in any industry?

The answer is a resounding no. Indeed, there is actually some evidence that companies that try to artificially force up the number of female workers to achieve gender parity in representation and pay are actually less profitable and underperform the market. This is something economics would predict if labour markets were previously efficient, and artificial external forces were imposed on them.

Therefore, an economics frame of reference would provide compelling evidence that there is no labour market pricing inefficiency - that women are paid less than men because, on average, they are less productive and/or on average are working in less productive industries or in industries where labour across both genders has less bargaining power. Importantly, this perspective in no way implies that this outcome is fair. Market outcomes are amoral. This point will be discussed later.

An additional datapoint comes from the pornography industry. Female actresses, on average, make about 2-5x as much as male actors in the pornography industry. That statistic will probably surprise no one, but it nevertheless serves to make the economics perspective quite emphatically - that labour markets are functioning efficiently and are responding to market forces, rather than being dictated by ideology or discrimination. Women are paid much more in this industry because the demand and supply balance of willing participants is heavily skewed in women's favour (talking purely from a dry economics perspective - I am not offering any opinion on the moral aspects of the industry).

#3 The evolutionary psychology/biology perspective

Yet a final perspective on the matter can be found in the evolutionary psychology and biology departments, and particularly the perspective offered by the dynamics of sexual selection.

Evolutionary psychologists believe that our emotional circuitry, including our sexual proclivities, have been evolved in the same manner as our physical characteristics, such as eyesight; the ability to speak and hear; our physical dexterity; and our bipedal locomotion: namely they exist because they have - over eons - promoted our survival and replication. We are viscerally afraid of heights, snakes, and spiders for a reason - these things can kill us. Humans that displayed a healthy fear of these things in the past were more likely to survive and pass on their genes. Similarly, men and women's sexual appetites compel them towards sexual behaviour that was most likely - over our long evolutionary history - to successfully pass their genes on to the next generation.

Importantly, this process is both amoral (see the discussion further below), and also at times irrational. We might still feel a fear of heights even when we know logically that we are merely standing on a very safe glass-bottomed floor in a sky tower. Our emotional circuitry is hardwired, and often defies the logical portion of our brain. We crave sugar because over our evolutionary past, calories were scarce, and death by starvation was a much graver risk than death by diabetes. We may logically know that we should not eat sugar, but we are emotionally compelled to do so and so often eat sugar anyway. Similarly, emotions sometimes compel humans to engage in sexual behaviour that is self destructive in nature and not necessarily in their long term interests. There are many examples of this in the popular press at the moment, not to mention scandals past (e.g. Tiger Woods and Bill Clinton).

The evolutionary psychology perspective will predict that men are more likely to work harder and earn more money than women because they have more of an evolutionary reproductive incentive to do so. This can be easily determined from a study of biology across a range of species, as well as a simple reflection on the realities of the sexual behaviour and preferences of people in the real world, as well as an examination of cross-cultural statistical evidence. Women across all cultures exhibit a strong tendency to marry up and across social dominance hierarchies, on average marrying men 5yrs older than them and who make more money than them. Even high-income women, on average, marry men that make even more than them still. This is not statistical chance or gold diggery - it reflects hardwired evolutionary preferences.

It works like this: across almost all species, females are the sexual selectors. Males compete with one another to try to move up social dominance hierarchies, and women then generally select the men towards the top of those hierarchies, and prefer to mate with them. This pattern is clearly observable across species. Males higher up dominance hierarchies, and/or who demonstrate more traits of physical, social, or intellectual strength, and who have superior access to survival resources capable of aiding in a female's survival and the survival of her offspring (and who demonstrate a willingness to invest resources in that regard), are more attractive to women and more sexually successful. This process also facilitates the species' selection for the best genes, and so aids in the evolutionary process.

Human female eggs are much rarer than male sperm. Females produce approximately one egg per month, whereas men produce an average of 1,000 sperm per second. In addition, in the event of a pregnancy (bearing in mind that contraception is a modern phenomenon that was absent throughout our evolutionary history), women stand to bear a considerably larger and life changing investment than men - a 9mth pregnancy to start, which can take a heavy toll on the body, coupled with the subsequent need to care for an infant. By contrast - at least in primal days - the male party could simply walk away after a fairly trivial investment of physical energy in the deed itself.

Women's optimal reproductive strategy - i.e. the one most conducive to her successfully passing on her genes - is therefore find the best guy she is reasonably likely to find who is willing to commit resources to her and her offspring's protection and survival, and enter into a long term relationship with him. Somewhat more darkly, her optimal strategy is also to very selectively break this standard mold and cheat with a guy through a short term coupling only if the guy offers truly elite genes, and is of a caliber so high that she would never be able to get him to commit to her long term. This is why women will throw themselves at famous and highly attractive musicians, actors, and athletes, offering no-strings-attached sex, to the bemusement of many regular men. These men offer once-in-a-lifetime genetic opportunities. This is also why an estimated 10% or greater of children have a genetic father different to the paternal father. Women are incentivised to cheat, but only very very selectively.

For men, the optimal reproductive strategy is different. If given the opportunity (most men are not), it is to go for quantity rather than quality in something of a scatter-gun approach: sleep with as many women as possible, and hope some of your resultant offspring survive. Genghis Khan is perhaps the most successful exponent of this amoral reproductive strategy - it is estimated that as many as 1 in 200 people currently on the face of the planet is descendant from him.

For most regular men who are not of elite genetic quality and hence to whom the opportunity to sleep with large numbers of women does not exist, the best reproductive strategy is instead to find a women to enter into a long term relationship with, and aid in the protection and provision of one's relatively small brood of offspring, to maximise their chance of survival. However, it remains optimal for such men to cheat via short term liaisons with other women on the side if presented with the opportunity to do so, as from an amoral reproductive perspective, it is all upside, as it offers an increased probability of one successfully passing on their genes, with relatively limited investment or risk involved.

There is therefore a deep evolutionary logic to the fact that women care more about quality, and men about quantity (both are respectively scarce to the other), and behave differently in the sexual sphere on account of these different priorities. And these differences explains a lot of things. They explain why the market for short term sex is heavily skewed in women's favour (go out to any bar on Friday night and take a look). It explains why, although most attractive women could go out to a bar any night of the week and find someone willing to sleep with them extremely quickly - a situation most guys would kill for - that they mostly do not do so very often. It also explains why there is more of a market for female prostitution than male prostitution (if women desired short term non-committal sex more than men, it would be women paying men for sex). It also explains men's fairly widespread propensity to cheat if given ample opportunity to do so.

However - bringing the discussion back to the topic at hand - it also helps explain why men earn more money than women: it all boils down to evolutionary incentives. Because women prefer to mate with men towards the top of male dominance hierarchies, there is a much greater evolutionary reproductive reward for men successfully acquiring wealth, prestige, power, and fame, than there is for women. Women are the selectors, and men therefore have an incentive to work extremely hard to one-up other men and make themselves as acceptable to women as possible. This likely results - on average - in men working harder; making larger sacrifices; and taking more risks, in pursuit of opportunities to advance up the economic hierarchy. And in all spheres of life, it is readily observable that competition - with material prospective rewards - drives higher levels of performance. 

In short, the evolutionary psychology perspective predicts that men will acquire more positions of economic and political power in society simply because they have a much greater incentive to work hard in order to do so, and that this is likely the key reason for the persistence of the gender pay gap.*

So what should we conclude from the above?

Integrating the three perspectives, I would argue, tells us a lot more than focusing on merely one perspective, and perhaps more importantly, guards against the risk of reaching inaccurate judgments with a high degree of confidence we are right, when in fact we are wrong.

My view is there is likely important insights to be extracted from all three perspectives. The evolutionary psychology perspective partly explains why we see - on average - men in more high income positions, and the economics perspective provides strong evidence that widespread systemic labour market inefficiencies are not present.** However, as I'll explore below, the economics of childbirth also likely play a significant role, and the manner in which it does so is highly unfair.

This leads into the sociological perspective, which I believe also has something to offer. While I believe that many if not most of the strident conclusions deriving from the sociological school of thought are highly ideological in nature and in many important respects fundamentally flawed, the perspective implicitly rests on the correct underlying notion that there is an important degree of systemic unfairness implicit in many of the outcomes we see, and to the extent we feasibly can, we ought to try to correct them. This is explored below.

Efficient labour markets are not necessarily fair

It is also important to emphasise is that the processes described in #2 and #3 are amoral in nature, and work much like the law of the jungle. In the Tanzanian Serengati, much as in any other state of nature, the strong survive and dominate the weak, and the weak are killed or eaten. This is not fair. It is not moral. It is just a raw evolutionary process.

Unfettered market economics works in much the same way. Those with market power are strong and can dominate the weak, which lack such market power. This often leads to highly-unfair outcomes. It has lead to (much needed) regulation to control the level of monopoly power that can be allowed to accumulate, and/or to regulate monopoly power where it exists or ought to exist (e.g. in the case of a natural monopoly). Market forces can lead to market failure, such as a tragedy of the commons, where resources such as fisheries can be depleted by free market forces. Regulations to manage fisheries resources to optimise the level of catch and allow for natural replenishment are needed.

As it pertains to the present discussion, the most obvious source of unfairness is that women bear a uniquely high burden when its comes to childbirth. Unlike their male counterparts, women need to have children in the earlier part of life (generally ages 25-35yo), and need to take time off work during pregnancy during a uniquely-important part of their career development. My personal career took off in leaps and bounds during this age period, and was unobstructed by childcare requirements. That gave me an unfair advantage over women who needed to take time off to have children.

The impact of pregnancy on women's productivity can also discourage employers from employing and/or promoting women into key positions, if they are around childbearing age and are at high risk of needing to imminently take time off to bear or care for children. Hiring and training workers is expensive; replacing key employees is time consuming and disruptive; and paying maternity leave is also costly. These realities make employing such woman - all else being equal - a far riskier and expensive proposition for employers. A likely and predictable consequence is that the market would pay women less than men to reflect that risk, and women who took time off during a crucial phase in their career development would also suffer inevitable career setbacks, also impacting their lifetime earnings. The economic evidence suggests that labour markets are correctly pricing in these impacts and risks, so the outcome is economically efficient, but it is not fair.

In addition, sexual selection is also unfair. Some people are born better looking than others. It is not fair on less physically attractive women that men are more attracted to more physically attractive women than them, and that men cannot help it. Equivalently, it is not fair to short men that women are more attracted to men with a large stature, and that women cannot help it. There are very logical and understandable evolutionary reasons for why things are the way they are, but they are not fair.

One of the key philosophical differences between the right and the left, in my view, is the degree to which the above reality is identified and leveled with. We live in a society with ideals centered on equality and fairness. However, nature is anything but - individuals are unequal, and outcomes are highly unfair. Some individuals or species are stronger than others, and in nature, the strong crush the weak. Most often, those individual differences are natural/genetic rather than earned. The most ruthless and amoral often dominate the kind and moral. Good does not always prevail over evil. Nice guys often finish last. Some individuals are more naturally attractive to the opposite sex than others, and therefore enjoy superior but unearned access to more fulfilling relationships and sex lives. We are far from equal in this respect, and everyone knows it.

There is not a whole lot we can do to change our innate sexual preferences, although societies have developed various codes of morality and standards of acceptable sexual conduct over the years to try to dampen the excesses and potentially negative social consequences the unfettered expression of our raw sexual proclivities could lead to. This is arguably where sexually repressive or illiberal social norms originated from (along with the need for population control in an era without contraceptives). We also condemn as a society - rightly I believe - people excessively indulging their raw sexual whims - particularly when it leads to foreseeable negative consequences (such as cheating on one's spouse). Essentially what we are saying - as a society - is that we aspire to be better than our raw instincts, and that we ought to expect people to rise above them and control them.

The same is true with respect to tempering the excesses of a market system. There has been a longstanding and uneasy tension between the sometimes amoral brutality of completely free markets (that work in an analogous fashion to an amoral state of nature in the Serengeti), and the other ideals and values we hold as a society - the value of community; social cohesion; family; a sense of fairness and a willingness to care for the weak and less fortunate. For this reason, we feel - to greater or lesser extents over time as the political process vacillates - the need to temper free market excesses.

This is where labour market restrictions come in, such as including minimum wage rules; minimum holiday rules; and restrictions on the terms of severance. We recognise that the labour market is not just some other market, like the market for - say - soybeans. Labour markets impact people's lives and the other values we hold dear. Laying off people summarily in a downturn without compensation can pose significant hardship on families and communities, with significantly negative consequences. Excessive inequality can lead to a breakdown in social cohesion and a rise in crime. And raw free markets can lead to unsafe and exploitative work practices, including hazardous work environments; the exploitation of child labour; and the overworking of employees for unlivable wages. As a society, we have recognised that some limits need be placed on the raw functioning of market economics in the domain of labour. We cannot make society completely fair - that notional is fanciful - but we can at least try to make it slightly less unfair.

The same can be said of the unfairness deriving from the burden of childbirth on women, and the negative impact that has on their labour market value. Pushing employers to try to equalise, to the extent practicable, female participation in their workforce, is a good thing to the extent that it helps neutralise lingering aversion to hiring women on account of their potential need to take time off to have children. Pushing for greater pay parity and greater female promotion into senior roles is also likely to have the impact of at least partly counteracting some of the lost career opportunities women suffer as a result of spending time bearing children. And forcing employers to provide maternity leave also helps further militate against the adverse economic consequences of childbirth.

Provided they are not carried too far, these are all worthwhile initiatives, because as a society, we not only need to aspire to greater fairness to the extent we can, but each generation we also need to replenish the population, and rear a new, healthy, and productive generation. At worst, market economics can discourage childbirth altogether, and result in precipitous declines in fertility rates that threaten the very sustainability of society (both Japan and Italy now have fertility rates below 1.5 children per female - below the sustainability threshold of 2.1). This is another tragedy of the commons. The sociological perspective is wrong on a lot of things, but is right to highlight that unfairness exists, and that we should try to address it to the extent we feasibly can.


*Incidentally, the lack of interdisciplinary thinking, and in particular, a lack of understanding of evolutionary psychology, was a major reason why James Damore of Google was so misunderstood and treated so unfairly (see my blog post on the matter here). According to the social science SJW camp of mental models, Damore was speaking sacrilege. However, according to the evolutionary psychology set of mental models, Damore was not being controversial at all - he was speaking well-accepted scientifically-documented truths that are the consensus view of evolutionary biologists. That doing so is now a fireable offence in the US in truly alarming.

**It is important to emphasise that these conclusion refer to broad population-level averages, or overall tendencies. Individual situations always have the capacity to be highly idiosyncratic and out of step with those averages. Labour markets are not perfectly efficient. And some women are as highly motivated to achieve high levels performance as men, while many individual men are lazy. Individual women have achieved extremely high levels of performance in many fields, and will continue to do so, and many women out-earn men. It is important to differentiate between generalising broad patterns of outcomes, and individual outcomes.