The Execution-Certainty Wedge: How Measurement Choices Rank Housing-Access Instruments

When sellers prefer offers that are more likely to close, cash buyers can win homes that financed buyers value more. This paper uses an agent-based housing market to compare four instruments that change who wins: a per-entity acquisition cap, a fair-share quota, a constant charge on institutional offers and a duty that rises with local contention. The ranking of these instruments depends on how access and efficiency are measured. An efficiency benchmark computed over the buyers still waiting moves with the policy, and for offer discounts between about 5% and 12% it reverses the sign of the duty's effect. Access measured over the whole run understates each instrument's effect on a cohort followed to purchase or exit, by factors of about 1.4 to 2.5, and ranks the instruments differently. Tuned to the same full-run access gain, a constant charge preserves more value than the contention duty; tuned to the same cohort gain, the duty keeps slightly more value at the ten-point target. At low institutional participation a fair-share quota that resets each period cannot reach small access targets, and letting its allowances carry forward removes that limit. New York City evidence does not support the model's premise as a description of that market, because much of the observed cash discount reflects who pays cash. The results are conditional comparisons in a stylised model, not a calibrated policy evaluation. Version 6.1 adds Section 6.1, which runs the model at New York City proxy values. The offer discount is set to 1.5% from the company-cash price gap in Who Gets the House? version 3.4 (upper confidence endpoint 5.3%), and institutional participation approximates the 11.4% of 2016–2025 house purchases made by companies paying cash in the linked dataset (22.8% for all companies). At those values 9.8% of owner-occupiers leave without buying, against 41.9% at the paper's baseline. Removing the cash discount entirely lowers that only to 9.5%, so the certainty wedge explains little of the model's access shortfall at New York inputs. The quota has no measured effect, the duty and the constant charge each improve access by about 0.2 points, and the cap improves it by 1.8 points at a 1.4% loss in realised value. Supply, arrivals, patience and valuations stay illustrative, and company status is only a proxy for certain institutional buyers. The rest of the paper is unchanged from version 6.0. Version 6.0 shortened the paper to about 7,700 words, organised around how measurement choices rank the instruments, without changing any result. It added a two-period example showing why a benchmark computed over the waiting buyers moves with the policy, and bootstrap intervals for the break-even offer discounts: 11.7% (95% interval 11.5–11.8) at the baseline and 15.9% (15.6–16.1) at the lowest institutional share. The extended version, with the full robustness exercises, the discussion of implementation and legal form, and the complete replication record, remains available as version 5.1 (doi:10.5281/zenodo.22977667) and is included in the replication archive. Files: the manuscript PDF; WEDGE_v6_1.zip, the replication package, which regenerates every result in both versions, including the New York City runs and their 67 text checks; and revision_notes.md. For details of changes between versions, see revision_notes.md.