AI for economists: data modeling, economic outlook reports and policy synthesis

If you’re an economist or economic analyst, a good chunk of your week goes to tasks that don’t require your best thinking: cleaning a data series, reformatting a table, summarizing a forty-page central bank report, writing the same outlook report with new figures. The sharp analysis, the part that truly adds value, gets left for the end, when you’re already tired. Artificial intelligence can handle much of that groundwork and give you back time to think.

Not to replace your judgment. To take away the mechanical part and leave you the analysis, which is where your value is.

What AI can do with your data

Think of AI as an assistant you hand the raw material to (a series, a report, a table) and it returns something nearly ready to review. In economic work, a modern AI can:

• Clean and organize a data series and explain what it did
• Compute growth rates, changes and averages without fighting formulas
• Summarize a long report (from a central bank, the IMF, a ministry) into its key points
• Draft an outlook report from your figures
• Synthesize an economic policy measure into plain language
• Suggest which charts best tell the story of your data

What used to take you an afternoon can now take minutes.

A real example: from the series to the report

Imagine you have the month’s inflation and activity figures in a spreadsheet. The flow is this simple:

1. You give it the data. Paste the table or upload the file to an AI like Claude or ChatGPT.
2. You tell it what you want. For example: “Compute the monthly and year-over-year change, note whether the trend is accelerating or slowing, and compare it to last quarter”.
3. You get the draft. The AI returns the calculations with clear text describing what’s happening.
4. You ask for the charts. “Give me the formulas or step by step to chart this in Excel, or the Python code if you prefer”.

You review every number, correct the analytical frame and sign the report. AI wrote the draft; the interpretation is yours.

Outlook reports without starting from zero

This is where you save a lot of time. Instead of staring at a blank page, give the AI your figures and notes and ask for a first draft with a clear structure: context, period data, reading and risks. Ask it to keep a technical, sober tone, without extra adjectives. In minutes you have a solid skeleton that you refine with your reading of the context, instead of writing it all from scratch.

Synthesizing economic policy in plain language

Another task AI is grateful for: translating. Hand it the text of a measure (a tax reform, an interest rate decision, a statement) and ask it to summarize it in three paragraphs for a non-expert audience, or to build a table of “what changes, who it affects, starting when”. It’s material ready for a note, a presentation or a client who doesn’t speak the technical language.

The important part: verify every number

AI is incredibly fast, but it’s not an economist and it can be badly wrong. It can make up a figure, misread a series or cite data that doesn’t exist (what’s called a “hallucination”). In economics, one wrong number ruins an entire analysis. So:

• Always verify every figure against the official source before publishing. This is not negotiable.
• Don’t trust calculations blindly: ask it to show the step by step and review it.
• Protect confidential data from clients or institutions; don’t upload it without confirming the tool’s privacy policies.
• Use it for drafts, synthesis and organization, not as your final source of truth nor as investment advice.

AI does the 80% of groundwork. The 20% that requires your economic judgment, the context and the responsibility of what you sign stays yours, and that’s where your value is.

Start small

You don’t have to automate your whole workflow tomorrow. Start with a single task: the report that weighs on you most or the series you struggle most to clean. Hand it to an AI this week, verify every number and see how much time it saves you.

If I, without being an economist, can build things with AI, you with your training and judgment can achieve so much more. You just have to start, and always verify.