The report

Your individual report (A004): the instructions, an example, and the checklist.

In short

• Deadline: Sunday 8 November 2026, 16:00. One PDF by e-mail to magnus.lodefalk@oru.se, subject NA310G: Assignment, file named with your personal code and -FINAL (for example K7Q2MX-FINAL.pdf).
• Resubmission: a report that needs a revision gets written feedback. The revised report is due Sunday 29 November 2026, 16:00, sent the same way.
• Length: at most 2,500 words, plus an appendix with your do-files, final log, the verification line and the AI declaration.
• Language: Swedish or English, and numbers by that language: in English 0.25 and 1,250, in Swedish 0,25 and 1 250. Do-files and logs stay as Stata writes them.
• Your question and data are personal: you download them with your code (page Your data, before Lab 1).
• Added controls: one or two of your own, motivated by theory or the article and tested jointly (testparm, Lab 2), are recommended for a VG but not required. The quality of the reasoning decides a VG.
• AI: to learn, never to do the analysis or write the report. See AI in the labs and the report.
• Your personal code, never your name, in the report, the file names, the do-files and the logs. Add your code at the end of each do-file’s name (2_sample_vars_K7Q2MX.do), so you can tell it from the lab’s version. The first lines of a log show your folder path: if it contains your name, delete that line in the PDF. Never use your personal identity number.
• Where you work: in a copy of your lab folder, editing the lab do-files to fit your question. Step by step, with what to delete; the occupation file is needed only for an instrument.

The instructions are the authoritative text: everything about the data, your research question, the template, the formalities, and what earns a pass and a VG.

Readings

The reading list from the instructions, with links. Starred (*): read carefully, and summarise one of them in your report. Hash (#): browse. The rest are optional. Every link is free to open; where the journal version sits behind a paywall, the link goes to the working-paper version.

• *Engberg, E., Görg, H., Lodefalk, M., Javed, F., Längkvist, M., Monteiro, N. P., Kyvik Nordås, H., Pulito, G., Schroeder, S. and Tang, A. (2024). AI Unboxed and Jobs: A Novel Measure and Firm-Level Evidence from Three Countries. IZA Discussion Paper No. 16717. PDF · IZA page. The source of the AI exposure measure, DAIOE, in your data.
• *Engberg, E., Hellsten, M., Javed, F., Lodefalk, M., Sabolová, R., Schroeder, S. and Tang, A. (2025). Artificial Intelligence, Hiring and Employment: Job Postings Evidence from Sweden. Applied Economics Letters, 33(16), 3138–3143. Working paper (Ratio WP 380), PDF · journal version.
• #Hellsten, M. (2025). Returns to AI Skills. Manuscript (Ratio Working Paper No. 378). PDF.
• #SCB. AI-användning i företag 2025 (the most recent version of the report earlier called Näringslivets användning av AI). Statistics Sweden. 2025 report · 2024 edition.
• Felten, E., Raj, M. and Seamans, R. (2021). Occupational, Industry, and Geographic Exposure to Artificial Intelligence: A Novel Dataset and Its Potential Uses. Strategic Management Journal, 42(12), 2195–2217. Article (open access).
• Engberg, E., Koch, M., Lodefalk, M. and Schroeder, S. (2025). Artificial Intelligence, Tasks, Skills and Wages: Worker-Level Evidence from Germany. Research Policy, 54(8), 105285. Article (open access).
• Babina, T., Fedyk, A., He, A. and Hodson, J. (2023). Firm Investments in Artificial Intelligence Technologies and Changes in Workforce Composition. NBER Working Paper No. 31325. NBER page.
• Acemoglu, D., Autor, D., Hazell, J. and Restrepo, P. (2022). Artificial Intelligence and Jobs: Evidence from Online Vacancies. Journal of Labor Economics, 40(S1), S293–S340. Working paper (NBER w28257) · journal version.
• Lodefalk, M. (2024). Artificiell intelligens och jobben. Ratio. A book in Swedish, in the library and bookshops. Ratio page.

Other sources. SCB statistics on firms’ AI use (Näringsverksamhet > IT-användning i företag > AI) and on the population’s use of generative AI (Levnadsförhållanden > Befolkningens it-användning), both in the Statistical Database; Eurostat on AI use in enterprises; the AI-Econ Lab’s papers and its AI Monitor.

Your reference list holds everything you cite in the text, in Harvard style; more sources are welcome.

An example report

A complete example report, on a question none of you has (the lab demo question, chosen to show the methods rather than to find anything about AI: do AI-adopting firms have higher financial expenses?), with notes on why each part is there: example_report.pdf. Use it to see what a complete report looks like; your question, sample and numbers are your own. Text taken from the example report is treated like any other copied text.

Before you hand in

Go through report_checklist.pdf in your project folder. It lists the things that most often sent a report back for a revision in earlier terms.