A concrete suggestion that the IA community might find useful is a forum to address some of the issues related to transparency, undocumented and poorly documented assumptions, and the different sets of subjective judgments made in integrated assessment models. This forum would provide the opportunity for those interested to understand the way individual models have been built. This would go some way towards addressing reproducibility and transparency issues, and also provide an informal, but structured mechanism for documenting assumptions and their implications.
We note that in any complex modeling effort, modelers will discover strange and unusual outcomes that may not appear to be reasonable. We have observed that in the climate modeling discipline for example, odd results and model ``fixes'' tend to remain within the purview of the individual modeling groups, and do not filter into the larger climate community. While this may be expedient in avoiding embarrassing explanations in the short term, it retards the development of the climate models in the longer term. Since such dirty laundry issues are also likely to loom large in the IA modeling context, it is important that the IA community find ways to facilitate communication about these issues. By openly discussing problems, solutions may emerge earlier, and outside communities will not have the feeling that the IA modeler's are presenting only the fresh linen.