Using alternative datasources like high-resolution satellite imagery and machine learning, we made the world’s first monthly subnational GDP dataset at a NUTS3 level.
Affiliated/ working with
Our data is granular, timely, frequently updated, and accurate.
We release GDP data before official sources, like the Office of National Statistics or Eurostat.
Our estimates are released with a one month lag. So data for January 2021 was released in March 2021.
This is significantly quicker than the 18 month lag of official sub-national datasets.
We provide new GDP estimates every month.
Our dataset covers monthly GDP estimates from January 2015 to the current month.
We ensure data accuracy by:
The RMSE of our data is 1%.
We take high-resolution satellite imagery and transform luminosity data into GDP estimates using machine learning.
Night light luminosity data taken from satellite imagery, has been used as a proxy for GDP, dating back to 2012 in the American Economic Review. We combine traditional approaches with proprietary machine learning methods.
We are a team of doctoral/post-doctoral researchers from the London School of Economics, partnered with the European Space Agency BIC UK. We have published in peer-reviewed journals, designed policies for national governments, and are powered by bleeding-edge tech like Open AI’s GPT-3.
We use deep learning to produce accurate, granular GDP estimates.
We are partnering with the European Space Agency BIC UK to produce these estimates.
Our team is composed of doctoral and post-doctoral researchers from the Department of Economic Geography at the London School of Economics, the #1 ranked institution in Economic Geography in the world.
We have previously created computable general equilibrium models, macroeconomic forecasts (at the national and sub-national level), and conducted policy evaluations for NGOs, such as the Institute for State Effectiveness, research institutions like the London School of Economics, and government agencies such as the Afghan Ministry of Finance.
We provide our data in flexible formats, including:
Our data is conveniently formatted to allow easy combination with other datasets.
We have created computable general equilibrium models, macroeconomic forecasts (at the national and sub-national level), and conducted policy evaluations for NGOs, such as the Institute for State Effectiveness, research institutions like the London School of Economics, and government agencies like the Afghan Ministry of Finance. Some services we offer: