Firm microsimulation and VAT policy analysis

Firm microsimulation and VAT policy analysis

Vahid Ahmadi  ( PolicyEngine/Research Associate )  —  “Firm microsimulation and VAT policy analysis”
July 1, 2026, 0:00 am TBC TBC
Conference presentation

I am a Research Associate at PolicyEngine, a nonprofit that provides free, open-source software to compute the impact of public policy in the US and UK. Previously, I served as a researcher at the London School of Economics. My work focuses on microsimulation, economic modelling, and public policy analysis, particularly the UK tax and benefit system.

In this session, I’m going to introduce a firm microsimulation methodology for policy analysis. This framework generates synthetic firm populations calibrated to official statistics through multi-objective optimisation, matching distributional targets across turnover, sector, and employment from multiple administrative data. We then layer behavioural responses onto this synthetic population using bunching estimation methods from Saez (2010), Kleven and Waseem (2013), and Liu and Lockwood (2015). We demonstrate the methodology through UK VAT threshold analysis: firms near the registration threshold face incentives to stay just below it, creating bunching in the turnover distribution; we estimate how strongly firms respond by comparing the observed distribution to a no-bunching counterfactual, then simulate how firms would redistribute under alternative policy designs. The session covers synthetic data generation, PyTorch-based calibration, behavioural response estimation, and policy simulation using PolicyEngine’s open-source Python-based platform.

While household microsimulation is widely used, the absence of firm microdata results in firm-level microsimulation being less common, despite businesses being central to tax policy. Developing firm-level microsimulation opens new research directions, particularly when linked to household models, enabling analysis of how business taxes pass through to consumer prices, how firm-level policy changes affect employment and wages, and how supply-side interventions are distributed across income groups.

One direct application is VAT threshold analysis: the UK requires businesses to register for VAT and charge 20 percent on sales once annual turnover exceeds £90,000, and there is active policy debate about raising this threshold. Firm microsimulation enables analysis of different registration thresholds and smoothed phase-ins on revenues and firm counts, questions relevant to the economic growth literature. We benchmark static results against HMRCs published projections for raising the threshold from £85,000 to £90,000, achieving approximately 90 percent alignment with official costings.