Time or Resources? For Allocare, It Depends on the Environment

– by Elic Weitzel, Kurt Wilson, & Rich Sosis (Image credit: John Shaver)

Alloparental care—investment in offspring that are not one’s own—takes many diverse forms. Alloparenting can look like an older sister watching over a younger brother while their parents are out. It can look like an aunt or uncle helping to pay for their nieces’ and nephews’ college education. It can look like a grandmother providing her grandchildren with food that she collected, or it can look like a teenage next-door neighbor being paid to babysit for an evening.

These various types of allocare all accomplish a key outcome, providing for another’s offspring, but they each have different costs and benefits. Some involve directly provisioning children with resources of some sort, such as food or money that once given cannot readily be shared or regained. Giving away resources means that the alloparent had to first pay the costs of obtaining the resource and second the cost of not consuming or using it themselves. Other alloparenting activities involve expenditures of time rather than resources, such as supervising or playing with children. These forms of care do not require material resources to be transferred, but the opportunity costs of spending time in allocare can be consequential. Time spent supervising a group of children is time not spent doing anything else.

In thinking about both different forms of care, we theorized that care could be divided up roughly into two types. The first is care which entails an additive cost structure in which an alloparent must pay an equal cost for each additional child they care for. These children receive the benefits of care directly in proportion to the costs paid by the alloparent: a 1:1 ratio. Generally speaking, this most commonly reflects resource investment. For example, when a wealthy aunt pays for $10,000 worth of their nephew’s college education, that is $10,000 the aunt no longer has. The second type of care involves a declining marginal cost structure in which each additional child cared for costs less than the previous (at least up to a point). This is generally exemplified by time investments, such as when an older child cares for their younger sibling. If they were to care for two younger siblings simultaneously, they would certainly incur additional costs but not twice the costs of supervising just one. The second sibling can be supervised more cheaply since the older sibling is already spending the time watching the first child. In contrast, if our hypothetical wealthy aunt were to cover two nephews’ college tuitions, she must fully pay the same amount for the second.

Children require some blend of time and resources invested in them to survive and thrive, but presumably there exist situations to which each of these types of allocare is better suited. To explore this, we constructed a type of computer simulation called an agent-based model. An agent-based model functions like a video game with many different characters, all programed to do certain things. The outcome of the interaction of all the different agents making their own, simple, decisions becomes the focus of study – because these interactions can cause quite complex phenomena to emerge. In our model, we designed four types of characters, or agents, that each behave differently, following the strategy and payoff categories of alloparenting. One type does not alloparent at all, but only provides care to their own children with an additive cost structure, or resource investment. The second type also only provides care to their own children but does so in line with a time investment strategy characterized by declining marginal costs. The third and fourth types of agents both parent and alloparent, providing care characterized by additive and declining marginal cost structures, respectively.

In our model, we gave each of these agents a non-specific “currency,” which can be thought of either as time or resources, and let them go about their lives. These agents age, reproduce, care for children, and die. As they do, we track how much currency they keep, how much they spend, who they share that currency with, and—importantly—how many times they reproduce and how many of their children reach reproductive maturity. We evaluated the evolutionary success of each care strategy by looking at the number of children born to members of that agent type who survive to adulthood themselves.

What we found was that the cost structures of allocare matter differently in different socioecological situations. When we provided agents with abundant currency (think a resource rich environment) the two allocare strategies outperformed the two parenting-only strategies, producing an average of 4.75 surviving offspring versus the parenting-only strategy’s average of 3.75. Agents who provided allocare performed equivalently well in these simulations regardless of whether they paid additive or declining marginal costs. In other words, when resources are abundant the type of care didn’t matter, only the fact that they were providing some form of allocare. However, when we reduced the amount of currency we provided to agents, simulating contexts of scarcity, only agents who provided allocare with declining marginal costs (time investment) did well while the other three types struggled to effectively birth and raise children. This is because time-investment allocare, with its declining marginal cost structure, is characterized by what is known as an economy of scale: an economic situation in which scaling up production allows you to lower your costs. By providing only care that has declining marginal costs, alloparents are able to gain evolutionary fitness benefits by producing more surviving offspring. When an agent alloparents for someone else, or someone else alloparents their children, neither is paying a 1:1 ratio of resource to child but saving some resources to better ensure greater survival.

While this model is necessarily less complex than the real-world, the results from this set of simulations allow us to make several predictions about real-world alloparenting. Based on these insights from our model, we expect that time investments—watching, supervising, or teaching children—will be more common in socioecological contexts of scarcity. In contrast, contexts of abundant time and resources might lead to more varied manifestations of allocare. We also predict that time-investment allocare with declining marginal costs will be more universal than resource-investment allocare with additive costs as it is adaptive over a wider set of socioecological conditions. More work is needed, but available ethnographic observations appear to support this prediction, with many anthropologists noting that time-investment care, such as children’s playgroups, are particularly common across societies and ecological contexts.

There remains much work to be done investigating the costs and benefits of different forms of allocare. Our model is, after all, only a model. Real-world anthropological data must now be marshalled to evaluate the predictions we outline. Future work can also adjust our model in various ways to explore related questions, such as the reality that children often require both forms of allocare or require different forms at different points in their development. As we move forward, combining the power of computational modeling with ethnographic observation provides a uniquely powerful approach to testing complex theoretical ideas that will better help us explain the commonalities and differences in alloparenting strategies across cultures.

Read the original article: Weitzel, E.M., Wilson, K.M., Spake, L., Schaffnit, S.B., Lynch, R., Sear, R., Shaver, J.H., Shenk, M.K., & Sosis, R. (2024). Cost structures and socioecological conditions impact the fitness outcomes of human alloparental care in agent-based model simulations. Evolution & Human Behavior, 45(5), 106613.