Flex Index moves to LSE and IE Madrid and becomes a free open research repository
Flex Index, the dataset tracking which employers offer which flexible working arrangements, announced on 25 August 2026 that its stewardship is transferring to academia. Professor Prithwiraj (Raj) Choudhury at the London School of Economics takes it on, working with Assistant Professor Marco Minervini at IE University in Madrid. The dataset becomes a free, university-hosted open repository, with quarterly reports published and the underlying data made public.
The first refresh under the new stewards is due on 6 October 2026, alongside a Remote and Distributed Research Initiative conference at LSE.
Why this matters
Flex Index has been the most detailed public record of employer flexibility policy, and until now it has been a commercial product with limited access to the raw data. Making it free and open changes what journalists, researchers, jobseekers and negotiating employees can do with it. If you are assessing whether an employer’s stated hybrid policy matches what comparable firms in its sector actually offer, a public dataset lets you check rather than guess. Mark 6 October as the date the first open version appears.
Context
The move also shifts the centre of gravity for this research towards Europe. Flexibility data has been dominated by United States sources and United States samples, which has made it awkward to apply to European labour markets where statutory rights, collective agreements and works councils shape arrangements far more than they do in the US. Hosting at LSE and IE Madrid puts two European institutions at the centre of it. That matters in a year when European evidence has been thin relative to the volume of commentary, and when national datasets such as Ireland’s Labour Force Survey and Eurofound’s European Working Conditions Survey have been the only reliable anchors.
What to watch
The 6 October release and the accompanying LSE conference. Whether the open dataset expands its European employer coverage, which has historically been the weaker part of the sample, is the question worth returning to.