In my recent post reviewing Paul Frijters and Gigi Foster’s An Economic Theory of Greed, Love, Groups, and Networks, I flagged that they considered their new theoretical contribution to be “the love principle”. In this post, I want to pull the idea apart. I’m not going to offer a perfect alternative to the love principle, but I hope that by giving it a good going over I can possibly understand it better.
Frijters and Foster state the love principle as follows:
Love derives from the attempt of the unconscious mind to bargain with something that is believed to be capable of fulfilling desires and that is perceived to be too powerful to be possessed by direct means.
The starting point for how the love principle works is a person who believes that an entity can deliver to them something that they want. This entity could be a family member, a stranger or an abstraction such as a god. The person then considers whether they can take what they want from the entity (i.e. they can dominate it) and if so, they take it (greed).
If they cannot take what they want through dominating the entity, their unconscious mind seeks to bargain with this entity (now the love object) in the hope that emotional investment will lead to them getting what they want. Frijters and Foster liken this response to (among other things) submission, although it is a more complete form of submission as there is an internal change in self-image. The person becomes partially “one” with their love object. They internalise the wellbeing of the love object into their own wellbeing and are willing to make transfers toward this love object.
Given that the love-greed response depends on whether a person can dominate an entity, the trigger for love is the nature of the power relationship. People will tend to love entities more powerful than themselves. This might appear to create difficulties where two people love each other, but Frijters and Foster try to make it work. As an example, they suggest that a young child wields power over their parents before birth through the costs that the mother must bear (hence, fathers are snared by children to a lesser degree). Conversely, children love their parents as they are initially helpless, and do not realise that they will have power later on.
As I foreshadowed in the last post, I am not a huge fan of the love principle for a few different reasons. The first is that their description of the agent becoming “one” with the love object is, in one particular case, more accurate than they suggest. Consequently, this groups together what I would argue are several different phenomena. The model that Frijters and Foster present at the end of the book is a useful tool to illustrate this point.
In this model, an agent lives for two periods and seeks to maximise his utility in both periods of his life. This completely naive individual (incorrectly) believes that an emotional investment in a love object will yield a more consumption in the first period. What the agent does not realise is that in the second period the agent’s utility function changes to incorporate the love object. Frijters and Foster do this by making the agent positively weight the love object’s utility in the agent’s own utility function, with the level of incorporation of the love object’s utility an increasing function of the agent’s initial level of emotional investment.
As consumption by the love object in the second period increases the agent’s utility, the agent transfers consumption goods to the love object. The net result of this action is that the agent consumes the same amount in the first period irrespective of their level of emotional investment, as the emotional investment does not actually increase their consumption. The agent consumes less in the second period due to their transfer to the love object. Their utility is higher in the second period, however, as they have now gained utility from the love object’s utility.
From an evolutionary standpoint, I’m not sure how this system could evolve as an agent that does not love would consume more. There would need to be some other advantage to love not incorporated into the model. Frijters and Foster hint at a few possibilities, such as confidence and signalling, but for some of the examples they consider, there is an important option – kin selection. It is easy to incorporate kin selection into the model, as the weighting given to the love object by the agent would be the degree of relatedness. This turns the utility equation in the second period into a modified form of Hamilton’s rule, with the asymmetry of transfers between parent and child based on the lower reproductive value of the parents.
By making the love object kin, the love object would be part of the agent from the perspective of the relevant unit of selection, the gene. To the extent there is any internalisation going on due to the triggers or primes that lead to children and parents loving each other, evolution shapes it so that people internalise kin according to their relatedness.
This approach creates a simpler explanation for some of the love relationships Frijters and Foster describe. Children and parents love each other as they share genes. Mothers love more than fathers as the mother is more sure that the child is hers, and a father can produce many more love objects in a lifetime to which he can give his love. As Frijters and Foster note, this love of kin is not automatic and involves priming before and after birth. They point to the power relationship as triggering this love, and suggest that children do not love machines because machines have no power. But there are alternative explanations to this, such as having a type of object (e.g. a human face) that they are susceptible to be primed with.
Of course, this approach does not extend to the cases where two non-kin love each other, which relates to the problem that I raised above. Love of kin has different evolutionary forces shaping it to love of partners, groups or gods. As a result, I’m not sure they should be grouped together, even if there are some common neurological or biological factors.
For these non-kin cases I take a different reading of the process leading to “love”. I like the build up to the love principle – decide what you want, see if you can take it, take it if you can. But I lose them at the point where the person decides they cannot dominate the love object. I would state the process as being that after someone assesses that they cannot get what they want by domination, they come up with a different strategy. Those strategies are complex and varied, but many of them are not love in the sense felt between kin, or even loyalty as it is usually defined (Frijters and Foster group love and loyalty together). Accordingly, many of the so-called battles between love and greed described in the book are simply assessments of the costs and benefits of different actions under varying constraints.
Consider this in the context of another example they give, a reasonably well matched couple. If each has what the other wants but they can’t simply take it, love will flow both ways. Social conditions may affect that power balance, such as the social stigma associated with one of the couple receiving services from elsewhere. There may be similar neurological and biochemical processes associated with this form of love to those that occur with children. But the relationship is different – reciprocal rather than becoming one. An indication of the difference is the likelihood that love between a couple will fade if conditions change, such as the male gaining more wealth and resources, which is markedly different to the usual case with children.
I’ve never come across a completely convincing evolutionary explanation for love outside of kin relationships, but lean towards love being a bootstrapped evolutionary solution to providing a commitment device. As argued by Robert Frank, love results in behaviour that honestly signals commitment to the other party. Or, my generally preferred explanation, it simply focuses the loving parties on the actions that will maximise their reproductive success.
Although the evolutionary biology approach doesn’t offer a perfect answer, it does work for the out-of-sample predictions. Power is an aphrodisiac, as a woman wants a powerful man’s genes and resources. As a person becomes more powerful, love fades as they simply have more opportunity or resources. Parental investment theory provides a basis for sex differences in the response to power.
When we move to love in groups (which is probably closer to loyalty as traditionally defined), Frijters and Foster describe the love that the submissive members of the group feel. In their examination of the evolutionary benefits of the love program, they suggest that love prevents stress as it allows internalisation of the success of the dominant group members (although the Whitehall and other experiments suggest low ranked people still experience stress – and if stress was costly, why do people feel stress?). They also point to survival of a long period of subordination through avoiding possibly harmful conflict. But I am not sure that these strategies are loyalty or love. The submissive member may be implementing kleptogamy, where lower status males within the group try alternative ways of accessing females (or as John Maynard Smith termed it, the “sneaky f*&%er” strategy – and if someone knows an authoritative source giving John Maynard Smith credit for this phrase, please let me know). They may form smaller coalitions, such as banding with other beta males, with the resulting reverse-dominance hierarchy keeping the highest status male from taking too many spoils for himself, lest he be stabbed in the night or overthrown. These are strategic relations, with greed at their heart. We don’t need a love principle to explain the fact that a weak male chooses to avoid conflict in the short-term and bide his time.
The one element of the love principle that seems strongest to me (and for which this evolutionarily informed approach is weakest) is for love of abstractions and symbolic expenses, such as sacrifices to gods. There are potential explanations such as signalling, maladaptation, spandrel or strengthened bonds in a group through the shared abstraction, but they are not completely satisfying. The love concept is neat in that it directly addresses this phenomena, but I am not convinced.
As I flagged in the last post, when Frijters and Foster turn to the applications of their theories, I become more strongly of the view that what they are calling love is simply constrained greed. For example, they note the victory of love in the lack of trade barriers, but under the process described in the book, it is the greed of politicians and economists within their own groups that brought trade barriers down. They are responding to the internal incentives of the group, and I am not sure the economists are feeling much love. Rather, it seems that they are labelling positive results for groups as love, even if no-one specifically feels love.
This points comes out further in their view on the balance between love and greed over time. In the short-term, they suggest we decide with our heart. However, evolutionary factors mean that we have to effectively be greedy in the medium term to avoid being eliminated, so in the medium term greed wins. In the longer-term, they suggest that love is winning out, as if we look at the level of nations, history tends towards institutions that recognise greed but constrain it for the common good. Is this a subtle shift in what they mean by love? I’m happy to call this love, but it may not directly relate to the love principle, as it does not require that people have become one with any love objects, and could simply be the result of constrained greed (which is what their evolutionary analysis of the medium term would suggest).
So, to pull a rather long post together, I’m not convinced about the mechanism behind the love principle, that the full scope of love and loyalty covered by the principle should be grouped together, and whether many of the actions labelled as love are anything but constrained greed. I’m open to being convinced on many of these points, but I’m struggling to accept the whole package. Still, its a pretty neat idea to pull apart some interesting issues.