Aki Lehtinen proposes a modification to my account of invisible-hand consequences. My account of invisible-hand consequences excludes unintended social consequences that were brought about by actions of individuals who were intending to bring about social consequences. Lehtinen argues that “the problem with this condition is that it is too strong: it rules out cases in which we are inclined to say that a model provides an invisible-hand account even though it violates the condition.” Here is what he says in his introduction:
“N. Emrah Aydinonat presents an informative account of invisible-hand explanations in his recent book (Aydinonat 2008). The purpose of this comment is to propose a slight modification in order to accommodate some cases that his account, by virtue of being too strong, rules out as invisible-hand models.
I propose that Lehtinen’s (2006, 2007a, 2007b, 2008) model of counterbalancing strategic votes should be taken to provide an invisible-hand account. The models presented in the different papers are different in their details, but I will refer to them collectively as the ‘counterbalancing model’ because they are all based on the same counterbalancing mechanism. The consequences depicted in this model violate only one of Aydinonat’s conditions, and I believe it is better to slightly re-formulate his account than to conclude that the model does not count as an invisible-hand account. Given that it is at least a potential invisible-hand model, and has not previously been discussed in the literature, analysing its properties may have some independent relevance to students of economic and social-scientific methodology because models that truly provide invisible-hand accounts are rare.”
Lehtinen, Aki (forthcoming) “Intentions in Invisible-Hand Accounts”, Journal of Economic Methodology. You may read Aki Lehtinen’s commentary by clicking here!
In The Invisible Hand in Economics, I make a distinction between model-worlds and the real world. I think Aki Lehtinen may be conflating the intentions of real world individual voters (which may have individual and social intentions at the same time) with model-world voters (which only have intentions directed at the individual level). In the model-worlds of Lehtinen (2006, 2007ab, 2008) model-individuals engage in expected utility calculations which (it seems) has nothing to do with social intentions. Thus, the results produced in these models seems to satisfy the conditions of my invisible-hand consequences. However, Lehtinen’s point requires close attention. I hope to address this issue in the coming months.