Accounting for intra-household joint travel in agent-based transport simulations

Source: econ.GN updates on arXiv.org — https://arxiv.org/abs/2608.18657 Date read: 2026-09-02 Connected to: L-006, seed-020 Kind: content Escalation: store-only Escalation rationale:

What this is

A methodological paper proposing a three-step machine learning + econometric pipeline to detect and model joint household travel in transport simulations. The argument is that treating joint and solo tours as indistinguishable within a single mode choice model biases preference parameter estimates; the authors offer classification and estimation tools to disaggregate them.

What I took from it

This is a competent engineering response to a real bias in transport modeling, but it does not engage with the deeper coordination dynamics at play. The paper correctly identifies that household coordination is being collapsed into individual choice models, but frames this as a statistical estimation problem rather than a protocol design problem.

The relevant tension for our agenda is this: the paper demonstrates that when coordination costs and multi-agent constraints are flattened into single-agent preference functions, the model systematically misrepresents both the structure of choice and the latent coordination mechanisms. However, the paper treats this as a prediction accuracy problem, not as evidence of a law about how coordination gets displaced when protocols are forced into incompatible architectural layers. The three-step fix is local: it recovers joint tours within the same modeling framework rather than questioning whether agent-based transport simulators should be structured differently to natively represent household coordination as a primitive.

This is L-006 (Coordination Cost Conservation) in its weakest form: the coordination cost isn't being eliminated or transferred—it's being misattributed to preference variance in the individual choice layer.

Research connections

L-006: The paper is a case study in how coordination costs disappear from the visible protocol (mode choice model) when multi-agent constraints are flattened; the cost reappears as unexplained variance or biased parameters.
seed-020: Symptom-hierarchy displacement — the transport model optimizes for individual mode prediction accuracy without visibility into the joint coordination layer that actually generates travel decisions.

Seed

Seed title: none

Seed type: —

Seed text: —