Analysts & researchers
Labor market data for analysts and researchers
You know what regional labor market data can and cannot tell you. What you want is a clean, reconciled panel you can interrogate immediately, at the geography you actually work at — with every derived figure documented in enough detail that you can audit the method rather than take it on trust.
What it is used for
A panel that is already reconciled
Twelve years, national through county, with taxonomy revisions handled, control totals checked and suppression represented as suppression rather than silently imputed.
Full distributions, not summaries
The whole wage distribution rather than a median, employment by ownership sector rather than a total, demographic detail rather than a headline rate. If the source publishes it, you can get at it.
Methods you can audit
Every derived and modeled figure is documented in full — the damping on the forecasts, the z-scoring behind occupational similarity, the allocation rule on pipeline balance. Disagree with a choice and you can see exactly what to adjust for.
Custom geographies
Counties, metros, micropolitan areas, CSAs, or any set of counties you define as a region — with the same method applied consistently across all of them.
Get the data out
CSV on every table, and API access to the underlying panel on the Enterprise tier — so the platform can feed your own models rather than being a dead end.
Validated against independent authorities
Twenty-six automated checks compare served figures against independent published totals, and every check that allows a tolerance documents why.
In practice
Questions you can answer in a sitting.
- Is the county-level panel consistent across all twelve years?
- What exactly is modeled here, and by what method?
- Does this represent suppression honestly, or impute over it?
- Can I define my own multi-county region and apply the same method to it?
- Can I export this into my own analysis rather than working inside a dashboard?
The methodology page documents every derived and modeled figure in full. That is the level of disclosure this audience should expect, and rarely gets.
See it with your own region.
A 30-minute walkthrough on your data, not a canned demo. You keep whatever we build during it.