Haiti, we're back !
We were gone for much longer than anticipated but we're pleased to announce our mobility data pipeline for Haiti is now back and running again.
This means we are now able to provide again estimates of current population numbers and population movements across Haiti, every month.
Data are now available from January 2020 to February 2026.
What happened?
From June 2025, the Flowminder-Digicel pipeline had been interrupted. This means that no CDRs were being uploaded on our servers. We have since completed a data backfill from November 2025 to date but we do have a data gap from June to October 2025 included.
What does it mean in practice?
This means that for now, we are not estimating anything unobserved for the months of June to October 2025: no population estimates for this period are available. We are planning to interpolate the data for those missing months but this will come in another monthly release.
As our method relies on the previous month to infer mobility for a given month, this also means that the mobility estimates we've produced for November are based on the changes between locations observed between May 2025 and November 2025 (rather than May-June; June-July etc..).
Since November 2025, we are back to producing the mobility estimates based on the consecutive months (November-December 2025; December 2025-January 2026 … etc).
For more information, please see our methodology documentation below:
What data are available and how to use this information?
Data available on the platform
The new estimates are now available on the Haiti Mobility Data Platform. The missing months do not show any data.
National overview report (crisis preparedness report)
Alongside this platform update, we are also releasing a new preparedness report covering the February 2025 - February 2026 period.
The preparedness report provides population estimates per communal section where there is sufficient coverage, per month, for the last 12 months, highlighting the areas where the population is unlikely to match planners data and expectations either because there's been an increasing or decreasing population compared to official projections.
Such information is important for risk mapping and contingency planning (population exposure), e.g. pre-positioning resources and shelters before the hurricane season, and generally for humanitarian needs assessment and service provision planning.
Acknowledgements
We thank Digicel Haiti for their ongoing commitment to data for good and for providing us with the pseudonymised CDR data.
We also want to acknowledge IADB and FAES thanks to whom our pipeline is funded for a period of 12 months.
Des questions ?
Please reach out to haiti@flowminder.org for any questions you may have on our data and how to use it. If you also have any feedback or are interested in supporting this work in the long-run, we’d love to hear from you.