CANVAS: A Canadian Behavioral Agent-Based Model
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
Economic models are valuable to central banks for conducting projection and policy analysis. The Bank of Canada’s current economic projection relies mainly on two complementary large-scale models—the Terms-of-Trade Economic Model (ToTEM) and the Large Empirical and Semi-structural model (LENS). However, introducing both household and firm differences at detailed levels and realistic behavior in these models can be challenging, both in theory and in practice. In this paper, we contribute to the development of Bank’s next generation of models with CANVAS, a Canadian behavioral agent-based model. We simulate individual behaviours of many different agents to provide an overall picture of the Canadian economy. CANVAS improves on earlier models in three ways: introducing household and firm differences at individual level, moving beyond rational expectations by incorporating realistic behaviours of real people and business, and modelling the Canadian production network. Finer details of difference (on demographic data like sex, age, occupation, and household balance sheets) can help policy-makers understand households’ consumption and employment decisions. By modelling the strategic price setting behaviour of individual firms with the lab and survey evidence, we also capture inflation dynamics through factors such as demand, supply, and expectation. The network structure in CANVAS connects agents’ different characteristics and their behaviour, putting it among the first class of macroeconomic agent-based models that can compete with benchmark models in out-of-sample forecasting performance. These features make CANVAS a distinct complement to the current models, with greater ability for forecasting and policy analysis.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.002 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.001 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.002 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it