Behavior bundles and relationship preferences: Initial thoughts.
Someone with a firmer grasp on calculus and microeconomics than me should help me with this idea I have for modeling relationship preferences based on “behavior bundles” and indifference curves. :P I have a general idea but I feel it could certainly be improved/expanded.
I’ve been asked by an anon if I could elaborate on this.
My micro class started the semester with consumer demand theory and and analyzing consumer preferences and behavior. A common way of modeling consumer preferences is by graph indifference curves. In a very simplified example, if there are only two goods in the world, like bread and peanut butter, that a consumer cares about, it’s easy to make a graph where quantity of good A is measured on the X axis and quantity of good B on the Y axis. Every point in this X,Y space represents a consumer bundle, or a combination of the two goods at set quantities. An indifference curve presents all consumer bundles that provide the same amount of utility/happiness/whathaveyou, such that the consumer would be indifferent between choosing any of the bundles on the curve.
Other things to know about indifference curves:
- Usually you see negative, convex indifference curves because people prefer to have more rather than fewer goods and to have a mix of goods rather than only one kind of goods. But there are plenty of exceptions.
- Because people have varying preferences, each graph only show the preferences of one consumer.
- There are an infinite number of indifference curves for any consumer preference graph, because there are infinite possible combinations of goods, but drawing a few curves will usually give you a good sense of things.
- Indifferences curves get numbered with labels like U1, U2, U3, … Un, numbered from lowest shown to highest shown utility. These labels only show relative utility, not absolute utility- you can tell whether a consumer would prefer bundle A or B or neither, but not by how much.
- Example graph below from Investopedia.
Okay, enough of that! While people tend to think about comparing physical goods, it’s very possible to compare intangible goods- time spent on work or leisure is a common one. The thought I had in class a week ago was to put behaviors on the axes and graph the preferences a person has to participate in specific behaviors in their relationship with another person (who I’ll call their partner for simplicity). I probably thought of this because I had been looking at the behavior checklist on aromantic aardvark and felt a need to display far more information than a checklist can hold.
Now, when we’re talking about behavior preferences in relationships, things get really hairy really fast. Even if you take just two behaviors like hugs and kisses, and draw curves to show your relative preference for different combinations of them in a relationship with your partner, any graph is going to have a very hard time representing what’s actually going on in your brain. There are lots of kinds of hugs and kisses. Long and short, public and private, mixed in with other activities or concentrated together. (And let’s not even get started on shades of behavioral concepts, like exclusivity.)
Even if you are able to lump behaviors into manageable categories because you are conveniently indifferent to these shades and variations and only care about the quantity/frequency of a given behavior, it can be really hard to assign utility to behaviors without accounting for the other person’s preferences. For example, I’ve been thinking a lot recently about a concept I’m currently calling dependent desire, when someone desires something because their partner desires it. If their partner had no interest in behavior x and didn’t bring it up at all, the first person wouldn’t think of it themselves (unless culturally pressured to or something) and they’re equally happy with the cases “partner likes x” and “partner doesn’t like x.” But GIVEN the case “partner likes x,” their preference is to x. Getting back to my model, if someone experiences a dependent desire to kiss their partner, then their indifference curves are going to look wildly different based on their partner’s desire to kiss them.
So yeah, information overload. What’s better at handling detail than graphs? Well, generally, you can see a lot more complexity in mathematic equations than in the graphs that result from them. In consumer demand theory, our principle concern is utility functions, which generate indifference curves and get written as U=f(x1,x2,…,xn), which shows that each x is a good (or behavior!) that contributes to a person’s utility. One common type, Cobb-Douglas functions, look like U = x2^α * x1^β and yield indifference curves of the standard negative convex shape, but models covering more than two behaviors and with dependent preferences would be much more complicated.
SO. At some point I want to look further into how much relationship information I can pack into utility functions, but my goal is actually less to build a working model for relationship preferences and more to use the attempt to build a model as a way to draw out pitfalls and complications and generate reflection and discussion. Does anyone already use visuals or math to model their relationships? Any suggestions for this model? Any corrections to my novice’s understanding of consumer demand? I’m all ears.