CEF 2026 Pre-conference Meeting

Computing in Economics and Finance (CEF) 2026

This page describes only the June 28 pre-conference on computational economics toolkits. It does not cover the main CEF 2026 scientific program (29 June–1 July) or the separate main-conference invited toolkit sessions; those are independent. Public drafts for the latter are in the SCE officers workspace on GitHub: toolkit sessions (overview), DDSL session, non-DDSL session.

• Location: Ca’ Foscari University of Venice, Italy, June 29 -- July 1, 2026

• Pre-conference date: Afternoon of June 28, 2026 (the day before the main conference); start time 14:00

• Venue: Meeting Room 1, Department of Economics, Ca’ Foscari University (Plesso A)

• Co-sponsors: Society for Computational Economics (SCE) and Econ-ARK (Sloan grant)

• Organizers: Chris Carroll (SCE President) and Alan Lujan (core developer, Econ-ARK)

• Purpose:

  • Kickoff of a contact group for researchers working on computational toolkits in economics

  • Create an agenda for building standards, interoperability and transparency in dynamic economic modeling

• Attendance vs. speaking: Attendance is open to any registered CEF participant. Speaking slots — the lightning survey and panel in Event 3, the DYNARE session in Event 2, and any other featured slots — are reserved for invited participants, with time at the end of the panel for floor comments from other attendees.

• Listener registration: A separate registration form for non-invited CEF attendees who wish to attend will be circulated by the local organizing team.

Schedule¶

The meeting begins at 14:00, with a coffee break at 15:45–16:15 (arranged with the venue); the program concludes by 18:00.

Participant status¶

Speaking slots are for invited participants. The table below is the public roster; detailed invitation tracking (message IDs, replies) is in preconference-responses.md in the SCE officers workspace.

Confirmed Attendees¶

Invited (response pending)¶

Context and motivation¶

Many other scientific and technical fields have converged on domain-specific standards and shared software tooling. That kind of convergence supports progress in a number of ways:

• by reducing the reinvention of wheels

• by making it easier for scholars to communicate

• by providing computational shoulders to stand on

• by fostering online communities

In many cases professional societies, standards bodies, or funding agencies have deliberately sponsored that convergence.

Event 1: Tower of Babel Problem in Economics (14:00–15:00, 60 minutes)¶

This session will provide an overview by the organizers of the opportunities to improve the computational landscape in economics, which seems to be well behind many of these other fields.

The main example in economics like the kinds of toolkits common in other fields is DYNARE, whose domain-specific-language (DSL) unified and accelerated DSGE modelling. But DYNARE covers only a slice of modern computational economics. An impressive number of other toolkits have been developed — heterogeneous-agent, sequence-space, deep learning, value function iteration, dynamic CGE, agent-based modeling, and more — and each has solved hard problems. They have not, however, cohered into anything like the DSGE ecosystem DYNARE produced (see, e.g.\ Volker Wieland’s Macroeconomic Model Data Base).

The breakneck pace of improvement in AI makes us optimistic that it is now possible to make major progress on this agenda. But AI will not solve the problem unaided and undirected. What is required is the development of a set of reference standards around which such an ecosystem can cohere.

A leading candidate for such a standard is an expanded DSL, built by extending the syntax of DYNARE, Dolo, or other existing tools, with the end goal of making it possible to rigorously and fully describe pretty much any model in economics. Such a canonical framework would not need universal adoption to do its work: even if individual researchers continued to use whichever toolkit suits them best, it would provide the figurative Rosetta Stone across toolkits — with AI as our Babel Fish — that the field currently lacks. Getting there will require sustained coordination among the kinds of people attending this conference.

Event 2: Lessons from DYNARE about What Is Needed (15:00–15:45, 45 minutes including audience questions)¶

Speaker: Sébastien Villemot (DYNARE / CEPREMAP)

The session draws on DYNARE’s experience as the most widely adopted toolkit in computational economics. The specific framing — what features of DYNARE’s trajectory bear most directly on the standards-and-interoperability agenda this meeting aims to launch — will be developed by Sébastien.

Coffee break (15:45–16:15)¶

Refreshments are provided by the venue.

Introducing the proposed working groups (16:15–16:25, 10 minutes)¶

Before the panel, the organizers will give a brief overview of the three proposed SCE Working Groups — (WG1) Language, Calculus & Formal Semantics; (WG2) Methods, Backends & Verification; and (WG3) AI Assistance, Reproducibility & Standards — so that attendees have them in mind during the discussion that follows. Scope details and the sign-up sheet are in the Working-group sign-up section; sign-ups are collected at the close of the meeting (Event 4).

Event 3: Lightning survey and audience discussion (16:25–17:45, 80 minutes)¶

Chair: Chris Carroll (SCE President; organizer)

• Lightning survey (16:25–17:25, ~60 minutes): Each invitee speaks for ~5 minutes

  • if invitee represents a toolkit, they react from the perspective of their toolkit

  • if invitee is not from a toolkit, they offer wisdom about requirements for success

  • ~12 participant speakers

• Audience questions and feedback (17:25–17:45, 20 minutes): open discussion, with comments and questions from attendees not on the panel.

Event 4: Working-group sign-up and kickoff (17:45–18:00, ~15 minutes)¶

The closing block makes the launch of a sustained process explicit. Building on the working-group overview given before the panel, the organizers will (i) circulate the working-group sign-up sheet described in the next section, (ii) nominate interim conveners and a meeting cadence, and (iii) identify 2-3 concrete near-term deliverables, including a follow-up working session at CEF 2027 and one or two other artefacts (candidates: a public catalog of toolkits and their interfaces; a shared notation document).

Working-group sign-up¶

The standards-and-interoperability effort launched here is organized as a set of standing Society for Computational Economics (SCE) Working Groups, each a forum that scholars may join, contribute to, and cite in grant proposals. The groups are organized by layer: WG1 supplies the semantic specification on which WG2 (numerical solution and verification) and WG3 (assisted use and institutional uptake) depend; WG2 in turn supplies declared numerical methods that WG3 can subject to automated robustness analysis.

Attendees are invited to sign up for one or more groups. The sign-up sheet will be circulated at the meeting (and online for those attending remotely); you may also indicate interest in serving as an interim convener.

A recurring cross-subfield integration aim runs through all three groups: letting models and components from different traditions — heterogeneous-agent macro, structural microeconometrics, computable general equilibrium (CGE), agent-based modeling (ABM), and applied IO — be expressed, shared, and coupled in one representation rather than re-implemented per tradition. Participants whose work spans subfields are encouraged to sign up under whichever group is the closest fit and flag the cross-cutting interest.

These groups carry the agenda forward between meetings, with an annual pre-conference meeting before each SCE (CEF) conference as the standing forum and a Working-Group Advisory Board drawn from the groups overseeing the standard.