Senior Data Scientist - Econometrics & Microdata

AG ECON, LLC
United States
2 days ago
Apply on arc.dev
Prepare application

Role details

Contract type
Permanent contract
Employment type
Part-time (≤ 32 hours)
Experience level
Expert
Experience required
5 years minimum
Working hours
Regular working hours
Languages
English, French, German
Job source

Tech stack

Artificial Intelligence Airflow Amazon Web Services Data Analysis Automation of Tests Code Review Data Warehousing Python (Programming Language) Machine Learning NoSQL Cloud Services Software Engineering
+5 more
SQL Databases Workflow Management Systems Git Information Technology Docker

Requirements

  • A Master’s or doctorate in economics, statistics, data science, mathematics, computer science or a related field, and at least 5 years of professional experience with messy, incompletely documented data from different sources.

  • A strong command of Python and its standard libraries for data analysis and statistical modelling. You are equally at home in SQL, ideally with solid familiarity with NoSQL databases as well.

  • You work the way software development does: with Git, code reviews and automated tests. You use AI tools where possible and sensible, but keep the understanding and the responsibility.

  • You know panel econometrics, know how to handle missing values, and validate model results out of sample. You use machine learning where it adds value.

  • You explain technical limits to economists and subject-matter requirements to data engineers - each in their own language.

  • You work independently and carefully, including with confidential data, and also like working in a team: people can rely on you, you make assumptions visible, clarify open points early, and give and accept feedback.

  • You speak English and either French or German.

Nice to have:

  • Experience with tax or register data, microsimulations, record linkage and/or company balance-sheet data.

  • Experience with cloud services (especially AWS), Docker, workflow orchestration (e.g. Airflow) or data warehouses.

About the company

BAK Economics is an independent economic research and consulting institute in Basel with around 25 employees from economics, development, data and AI. We work for the public sector, associations and companies, and we run our own data products.

Our projects are increasingly based on detailed microdata sets. These include, for example, tax data of individuals and legal entities, on which we build forecasting models, as well as company income statements and balance sheets, and patent data that we link to companies and regions.

For this we are looking for an experienced person who combines methodology and engineering: someone who discusses model choice and assumptions with our economists, and data models and pipelines with our data engineers. Do you know when fixed effects solve a problem and when they only hide it? Can you turn an economic question into a working prototype in a few days? And do you see datasets full of typos, inconsistent spellings and special cases as a challenge rather than a nuisance?

Then you are exactly right for us!

Your role

You own the models on our microdata in terms of content and methodology, from the raw data delivery to the delivered figure. You bring order to complex data, link different sources, and detect methodology changes and breaks in time series. You want to understand how data comes into being: which processes and rules lie behind a tax assessment or a balance sheet, and how they show up in the data. When something doesn’t add up, you stay on it until you know the cause.

On this basis you develop statistical and econometric models, produce forecasts and simulations, and examine, for example, how a tax reform affects a canton’s revenues. You test new ideas quickly: from a question raised by our economists you build a prototype in a short time, which shows whether an approach holds up.

You are the interface between economics and data engineering. You translate subject-matter requirements into clear specifications. What proves successful, you’ll bring into production together with our senior data engineers: they own infrastructure and pipelines, you own the methodology and deliver code they can take over - modular, tested and discussed in code review.

Alongside modelling, you think about how the correctness of data and results can be verified, for instance with plausibility checks, reconciliation against independent sources, and tests that run automatically with every new data delivery.

You question methods, spot implausible figures, and present the results to our clients clearly and traceably - including their foundations and limits. As a senior you shape our methodological standards, give colleagues feedback on models and code, and take responsibility for the modelling in client projects.

Many of our models are in long-term use - for instance in our patent analysis, our industry forecasts, our tax calculator, and our benchmarking and regional tools. That is why we place great value on your work remaining usable in the long run: cleanly documented, version-controlled in Git, and comprehensible and reproducible for others.

Apply for this position

This job is hosted externally. Click below to view the full posting and apply.

Apply on arc.dev
Prepare application

Good distractions

Loading talks and stories from around this role…