Micro –> Macro –> Micro, etc.

Kevin has a few threads regarding the effect that micro behaviors have when aggregated to macro behaviors:

It occurred to me as I was reading this Huffington Post article that there is a reverse-emergent dynamic that occurs when countries (often through their leaders) send signals to other countries through word and action. That is, if the actions of a group can be seen as emerging from the sum total of actions of its constituents, then it’s also true that the actions of the constituents are influenced by the information received at the group level.  Obama, speaking for the U.S., says to an Arab nation “we will treat you with dignity and respect if you treat us that way”, and this has an effect on individuals in that nation as to how they will behave individually.

In the past I’ve characterized the downward influence (from level 2 to level 1) as constraining autonomy, to which Kevin objects.  While I’m not sure that we’ve come to consensus on this particular sticky wicket, there’s clearly a connection between the superfoo thread and the macro/micro threads.

  • That’s one of my favorite things to think about right now. I hope it turns into a good dissertation after I finish this MA thingy… the creation of emergent structures in social systems and feedback between that structure and agent behavior.

  • kevindick

    I agree. One of the specific features of modern macro models that overcame my previous objections is that they treat expectations as endogenous–with feedback between the real performance and people’s expectations of performance.

    We definitely agree that such feedback occurs between us and the superfoo. I just don’t see this as constraining autonomy. I can still do whatever I want. I still respond to my environment. The environment is simply more endogenously determined than on the savanna.

  • Brady

    “then it’s also true that the actions of the constituents are influenced by the information received at the group level.”
    ^Did you mean ” received ‘from’ the group level”?

    Top-down feedback might constrain autonomy only if the bottom-level agent is for some reason constrained to be a part of the superfoo. For example, top-down feedback may require a change that, if not adhered to, would remove the lower-agent from superfoo-membership. If for some reason this excision is not a possibility for the lower-agent, then maybe it constrains autonomy. Otherwise, I don’t see it (I sided with most of Kevin’s points in your previous discussion).

    Doesn’t the superfoo depend on this top-down feedback for its very existence?

  • @Brady: I meant received “at” the group level (i.e. group to group) but then subsequently transmitted down to the individual level. In the case of a media interview, these two steps happen pretty much at once. But in a diplomacy meeting between heads of state, it is two separate steps, and of course information gets distorted as it travels down (after being transmitted across).

    I think you hit on a key issue in the debate between myself and Kevin. It might be the case that our difference has to do with the tacit ability of the individual to still be considered a member of the group or not. In other words, the constraints that I am referring to may not be absolute constraints — the individual may still be able to choose how they behave with maximal freedom — however some of those choices may either explicitly or implicitly remove the individual from the group and thus make them no longer a constituent part of the superfoo. In essence they become cancerous to the superfoo. The “health” of the superfoo is measured by the coherence/cooperation of the individuals, and thus for the superfoo to be maximally strong, the individuals must sacrifice some (but not all) of their autonomy.

  • Brady

    thanks for the clarification about at/from.

    I’m mostly on board with your last paragraph, and think that it clarifies your previous position. I think conditional restraints on an ‘ultimate’ sense of autonomy make sense, especially when tied to group membership. If you start with an assumption or given- group membership- then some choices are just not to be made.

    Two reservations: 1) I’m not sure I would consider the ex-superfoo individuals (we need a better term for a constituent of a superfoo) to be necessarily cancerous to the superfoo. I think that implies things about their membership in both that and other superfoo. Additionally, I think the cancer analogy implies something about the individual’s future interaction and causal relationship to that superfoo that I’m not sure needs to be included. I realize you were speaking in general terms- I’m just pointing out that the state of the post-superfoo individual may be important to consider in a discussion about autonomy.

    2) It seems reasonable to say that the individual also gains a degree of autonomy by membership in the superfoo. There are positive, dynamic benefits going both ways. And while the superfoo likely suffers a little from the fact that there is individual autonomy below it, it’s that autonomy that permits it to exist. Rather than maximally strong, I would say the superfoo is optimally strong, meaning that it if it was too strong (or maximally strong), the superfoo might overpower or constrain some of the autonomy at the lower level in such a way that it’s ‘health’ would suffer. So there are autonomy (or health) constraints going both ways.

  • kevindick

    For some reason, I thought Rafe believed one had no choice but to be a member of a superfoo. Wasn’t that the point of our sub-discussion about a superfoo having to offer a competitive set of advantages to its members? I still see shades of this belief in the “cancer” analogy.

    In any case, if we agree that there are multiple superfoos that people can join and leave at will, I am at peace with Rafe’s position. In every transaction, you give up something to get something. But it’s your choice. I’ve chosen to be a member of the Bay Area superfoo, which does prevent me from fully enjoying the benefits of other metropolitan area superfoos.

  • I believe there are/will be multiple superfoos, but only one Superfoo (by construction). So to the extent we are talking about groups where individuals can easily join or leave at will, then yes we are on the same page. We may have disagreement about what it means to leave or join. I don’t see it as a binary proposition. Where there is a lot of interdependence already established, leaving is less of an option. In this sense autonomy is constrained.

    And in speaking of The Superfoo, the more interdependent we all become, there will be some constraint on autonomy. There already is, and we are seeing the conflict between individual autonomy and group survival.