Economist
| Median wage (USD) | 124720 |
|---|---|
| Projected employment growth (%) | 5 |
| AI exposure (%) | 46 |
| Maximum exposure (%) | 58 |
| Source | eloundou |
| What AI changes in this occupation | Drafting the economic note, compiling and formatting series, producing tables and charts, estimating a standard econometric model, reviewing the literature and producing a first-draft forecast now take minutes with an assistant: the chain from raw data to published note — the one that filled most of a research economist's week — is the share being delegated first Management and clients now arrive with an analysis already produced by AI, which shifts demand towards checking the assumptions, the specification and the robustness of the conclusions; the economic briefing, the dashboard and the descriptive study are self-served inside organisations; and the economist is increasingly called in upstream, on framing the question and building the data, rather than downstream on the computation Saying what the figure does not prove, choosing the specification that actually answers the question, standing behind a public-policy recommendation before a committee, a legislative hearing or a court, and carrying responsibility for the published conclusion before the people who will decide remain human — it is the commitment to interpretation, in a job where two honest models give two different numbers, that makes the economist |
| Key skills | Microeconomics and macroeconomics · econometrics and causal inference · statistics and data handling (Python, R, SQL) · building and documenting datasets · cost-benefit analysis and public-policy evaluation · forecasting and modelling · checking specification and assumptions · economic writing for non-specialist decision-makers · public communication and testimony · generative AI tools for literature review, code and drafting |
| Courses to get ahead | Microeconomics Principles · Coursera (University of Illinois Urbana-Champaign) · 9 hours in total · 90 · The foundation AI does not replace: building an argument from incentives, opportunity costs and equilibria, then explaining why a counter-intuitive result holds. That is precisely what separates the economist from a producer of tables — an assistant writes the regression, it does not say whether the mechanism it measures is the right one. Understanding economic policymaking · Coursera (IE Business School) · 6 weeks, 2 to 3 hours a week · 86 · The side that stays human: reading a public policy, estimating its effects and arbitrating between conflicting objectives — growth, inflation, employment, redistribution. The course covers the part of the job where the economist commits to a recommendation before decision-makers — the part the ζ of 58% makes transformable without making it delegable. AI For Everyone · Coursera (DeepLearning.AI) · 4 weeks, 2 to 3 hours a week · 84 · For deciding what to delegate and what to keep: the course explains what generative AI can do, where it fails and how to frame a project that uses it. An economist whose compilation, code and first draft now take minutes needs to know what these tools actually produce — and what they cannot commit to. Financial Markets · Coursera (Yale University) · 7 weeks, 3 to 4 hours a week · 82 Quantitative Methods · Coursera (University of Amsterdam) · 8 weeks, 4 to 5 hours a week · 80 Game Theory · Coursera (Stanford Online) · 2 weeks at 10 hours a week · 78 |
| Occupations to pivot toward | Operations Research Analyst · 88940 · 12 · 63 Financial Analyst · 103570 · 7 · 50 Urban and Regional Planner · 89320 · 4 · 58 |
| Skills to acquire for the pivot | Eloundou measures the economist's exposure at 46% (human β) and up to 58% (human ζ): compiling and reporting data, studying series in one's field, assessing the socioeconomic impact of public policies, reviewing documents written by others and developing guidelines and forecasts are largely delegable, but the advice given to a company or an agency, the policy recommendation, explaining the impact to the public, published research and testifying before a committee remain human. Within the repository, 46 puts the economist 42nd of the ninety-four occupations: forty-one occupations have a higher exposure, fifty-two a lower one, none is tied, and its nearest neighbours are the postsecondary teacher (47) just above and the software developer (45) just below. Its ζ of 58 places it 54th, tied with the buyer (58), behind the pharmacy technician (62) and ahead of the nurse (57). The explanation is direct: the job produces a written deliverable — a note, a model, a forecast — that the machine now produces too, but it commits to an interpretation and a responsibility the machine does not commit to; that is why its GPT-4 self-rated series (β 50, ζ 100) is kept in expositionGpt4 and is not displayed, in line with the decision of 21 August 2026. BLS projects 5% growth in economist employment from 2025 to 2035, faster than the average for all occupations, with about 1,100 openings a year over the decade: the 18,600 jobs of 2025 would rise by 900. The driver cited is demand for quantitative analysis — organisations turn to economists to apply big-data analysis to pricing, advertising and other areas — and employment remains concentrated in the public sector: O*NET ties 52% of jobs to government, 23% to scientific and technical services and 10% to finance and insurance. The reading differs by vintage: O*NET, which carries the BLS 2024-2034 projections, rates growth as "slower than average (1% to 2%)" for 17,600 workers in 2024 and 900 projected openings, while the Occupational Outlook Handbook page published in August 2026 gives 5% for 2025-2035; both figures are cited here because both come from BLS, and the gap is a matter of projection period. The ζ of 58% says a little over half the job is at least partly transformable: what protects employment is not technical shelter, but the fact that the hiring decision rests on an interpretation a model cannot assume. The occupation's median salary is $124,720 a year, the May 2025 figure from BLS published on the Occupational Outlook Handbook page "Economists" — $59.96 an hour, with 10% of workers below $67,360 and 10% above $238,060. The same release shows the employer effect: $144,680 in the federal government, $141,010 in management, scientific and technical consulting services, $139,880 in scientific research and development services, $100,370 in local government and $78,080 in state government. For context, in the same May 2025 snapshot: all occupations stand at $50,980 and the social scientists and related workers group at $97,900; the three pivots on this card — operations research analyst ($88,940, +12%), financial analyst ($103,570, +7%) and urban and regional planner ($89,320, +4%) — each come from their own BLS page. The job usually requires a master's degree (50% of O*NET respondents), sometimes a doctorate (25%), rarely a bachelor's alone (21%); O*NET places it in "Job Zone Five: Extensive Preparation Needed" and does not flag it as a Bright Outlook occupation. The pay ceiling comes from high-level economic consulting, heading a research department or finance, not from the volume of notes produced. Every value on this card comes from a named source; only the task shares, derived from O*NET importance scores, and the course scores are editorial estimates, documented in _ponderation. |
| GPT-4 exposure — beta / zeta (%) | 50 · 100 |