How to apply for Economist, Scenarios & Modeling

About Windfall Trust

Windfall is a nonprofit that prepares society for economic disruption caused by transformative AI. It has no existing modeling infrastructure, so the person hired will build the quantitative capability from scratch. The work is remote and part-time, with potential to become full-time.

About the role

As the Economist, Scenarios & Modeling, you will own Windfall's quantitative modeling capability. The immediate priority is a scenario modeling platform that turns narrative AI futures into quantified economic parameters, propagates shocks through household microdata, and produces distributional results for policymakers. The role reports to the Director of Research and starts as soon as possible.

A typical day

A typical day might involve writing code to assemble household microdata, testing how a set of AI-driven economic shocks propagate through the data, and drafting a memo on distributional results for the Director of Research. You might also spend time choosing between modeling approaches, documenting methods for external review, or preparing to defend results to peer reviewers and government analysts.

Who Windfall Trust is looking for

• Has experience building economic models from scratch, especially scenario or microsimulation models that link macro shocks to household-level outcomes.
• Can write and direct code for data assembly, modeling, and visualization, and is comfortable defending methods to external economists and government analysts.
• Has range across quantitative methods, such as composite indices, reduced-form macro shock generators, fiscal tax modeling, or welfare system stress-testing, rather than deep specialization in one technique.
• Is self-directed and can choose methods, assemble data, and produce results without existing infrastructure or close supervision.

Tips for this application

• Show a concrete example of a model you built from scratch, including the question it answered, the data you used, and how you validated it.
• Explain how you would translate a narrative AI future into quantified economic parameters, since that is the immediate priority for this role.
• Describe your experience with household microdata and distributional analysis, because the platform must show who loses income, by how much, and what it costs the fiscal system.
• Address your ability to work part-time and remotely while owning a build role, and note your availability to start as soon as possible.
• If you have worked with policy audiences or peer reviewers, mention how you defended your modeling choices to them.

What to cover in your cover letter

['Your approach to building a scenario modeling platform that links AI-driven shocks to household microdata and distributional outcomes.', 'A specific project where you chose the methods, assembled the data, wrote the code, and defended the results to external experts.', 'Your range across quantitative methods, such as composite indices, macro shock generators, fiscal modeling, or welfare stress-testing.', 'Why you want to work at a nonprofit with no existing modeling infrastructure and can take ownership of the technical direction.']

Research before applying

• Read Windfall's published work at windfalltrust.org to understand its current research and how modeling would support it.
• Look for any public statements or papers on transformative AI and economic disruption to see the scenarios Windfall cares about.
• Check the Director of Research's background and publications to understand the technical standards and interests of your likely manager.
• Search for existing scenario models or microsimulation platforms used in AI economics to see what methods are common and where Windfall might fit.

Likely interview topics

Based on the job description, expect questions about:

• How would you design a scenario modeling platform that translates narrative AI futures into quantified economic parameters and propagates them through household microdata?
• Walk us through a time you built a quantitative model from scratch. What methods did you choose and why?
• How do you validate a model when there is no existing infrastructure or benchmark?
• How would you estimate the fiscal cost of income losses under different AI disruption scenarios?
• How do you decide which modeling approach fits a given question, for example composite indices versus reduced-form macro shock generators?

Common mistakes to avoid

• Presenting only academic models that were never used for policy decisions or defended to external analysts.
• Claiming expertise in every quantitative method without concrete examples of building models from scratch.
• Ignoring the part-time, remote, build-from-scratch nature of the role and applying as if it were a standard research position.