28/8/2019 ☼ Diversity ☼ Theory ☼ Innovation ☼ Adaptability ☼ Culture ☼ Cities
What are we talking about when we talk about diversity? Invariably we are talking about a group or population, since diversity definitionally requires difference within a plurality.
Most of the time, the conventional approach to diversity is thinking about simple differences in a group—but there are good reasons for thinking about diversity as 1) relative magnitude of difference between individuals within a group, 2) relative magnitudes of differences between subgroups of individuals within that group, and 3) multiple dimensions of difference within groups and their subgroups.
Thinking about diversity in this more complex way allows us to understand diversity better and thus promote it more effectively—and this is important because more diverse groups are both more innovative and more adaptable.
A population comprises an arbitrarily circumscribed set of individuals. Eight million people in London are a population; four people in a room are a population too.
The individuals in a population can be identical to or different from each other. If the four people in that room are identical quadruplets, they are a clonal population. If they are fraternal quadruplets, the members of that population would be technically genetically different from each other—as would be the case if they were are unrelated to each other (as most of the people in London are). The fraternal quadruplets are, technically, a diverse population.
The differences between individuals in a population can be large or small. The fraternal quadruplets are different from each other at a genetic level—they are technically a diverse population in the sense that they are not clones of each other. However, they are ultimately the genetic product of the same two parents. The inter-individual differences between identical quadruplets is zero; between fraternal quadruplets, non-zero but relatively small; between four unrelated individuals (children of eight unrelated parents), non-zero and relatively larger. When we consider diversity to be not just simple difference but the magnitude of difference between individuals, the identical quadruplets are undiverse, the fraternal quadruplets slightly diverse, the unrelated four individuals are much more diverse.
Relatedness and unrelatedness are themselves relative terms. A person whose ancestors are southern Chinese for many generations back through both parental lines is more closely related to another person of a similar ancestry than to a person whose parents’ ancestries are both north African many generations back. From a genetic perspective, the degree of relatedness between two individuals corresponds to the amount of shared or overlapping genetic information. The more shared information, the more related the individuals are even if they are not identical.
Variation in the relatedness of individuals in a population means that they can be grouped so individuals in a subgroup are different from each other but are relatively more related to each other than they are to individuals in other subgroups. Individuals in a north African subgroup are different from individuals in a south African subgroup—but they are less different from each other than they would be to individuals in a southern Chinese subgroup.
Populations can be diverse on two levels:
London is multilevel diverse. It is diverse in the first sense because it contains a large number of different people. It is also diverse in the second sense because it contains many subgroups of people. [Londoners] comprise [Londoners of southern Chinese ancestry], [Londoners of north African ancestry], [Londoners of Icelandic ancestry], and so on.
| City A | City B |
|---|---|
| north Japanese | north African |
| south Japanese | north Norwegian |
| east Japanese | south Chinese |
| north Chinese | Sardinian |
| south Chinese | west Brazilian |
City A and City B are similarly multilevel diverse, from a simple difference perspective. When the magnitude of difference is also considered, City A is less diverse than City B. City B contains more (genetic) information than City A, because it comprises subgroups that are more different from each other than City A’s subgroups.
Until now, the only dimension of difference I’ve discussed has been genetics. In real life, multiple dimensions of difference coexist. Genetically identical twins may have different jobs (professional difference), have gone to different schools (educational difference), and support different political parties (political difference). They may like different music, food, or fashion (all forms of taste difference). One may save more than the other (inter-temporal substitution difference) or enjoy dangerous sports where the other is averse to them (risk preference difference).
Diversity in a population should not be measured simply in terms of the magnitude of difference across levels when considering one dimension of difference—it should be about the magnitudes of difference across levels in multiple dimensions. The greater the magnitudes of multilevel, multilevel differences within a population, the more information (in multiple dimensions) the population contains.
Diversity is how information-rich a population is. A “true” diversity measure should conceptually enfold as many dimensions of information as possible at as many levels of analysis as possible. This is important because diversity-promoting policies based on simplistic, low-dimensional diversity measures are definitionally less effective than those based on complex, high-dimensional diversity measures which are closer to reality.
And promoting true diversity is vital for populations for at least two reasons, both related to the amount of information contained in diverse populations.
Diverse populations are more innovative. Innovation occurs through several mechanisms. One of the most important mechanisms is through recombination—combining the information contained in previously separate ideas/methods/objects to create new ones with new uses. The more information a population contains, the larger the pool of information from which recombinant innovations can emerge.
Diverse populations are more adaptable. Any discrete piece of information in a population is more or less adapted to particular environmental states. In a country where climate and geography make growing wheat cheap and easy, gluten-tolerant individuals perform better than gluten-intolerant individuals—a population of only gluten-tolerant individuals would be highly adapted to this environment. If the environment suddenly changes to make growing wheat both hard and expensive, while making rice (which is gluten-free) both easy and cheap, gluten-intolerant individuals will suddenly perform relatively better than gluten-tolerant individuals. If such an environmental change occurs, all else equal, a population comprising both gluten-tolerant and gluten-intolerant individuals will perform better than one with only gluten-tolerant individuals.
Information in a population represents a store of potentially useful responses to changes in the environment—and not just genetic information but also the other dimensions of information I alluded to above. When the population’s environment changes to one which is uncertain, it is important for the population to contain individuals who are not paralyzed by being exposed to uncertainty. When the environment changes to one where international exchange is easy and inexpensive, it is important to have individuals who are not xenophobic. The more unpredictably changeable the environment is, the more essential it is to have a population which contains more (and more different types of) information.