What estimated means
An estimated value is one BLS withheld to protect an employer and DTG produced under the constraints of everything BLS did publish around it.
The short answer
An estimate is bounded: the published parent total, the published sibling industries, the containing geographies and the published establishment count all limit what it can be, and DTG’s estimate sits inside those limits. A high reliability score means the limits are tight. A low score means the published record is thin there, and local knowledge, such as an employer you know opened or closed, should weigh more against the figure.
The detail
A zero in a published value is a measured zero. A very small estimate means the estimate is small, not that the industry is absent; an estimate below one displays as 0 on screen, the row details say “Estimate below one, shown as 0; the export carries the stored value”, and exports carry that stored value. Establishment counts are published even where employment is withheld, so establishments beside a small estimate indicate real but limited activity.
When counties are summed into an area cluster, the sum is marked estimated if any component was, and its reliability score is an employment-weighted average, so a large cluster can score close to 1 while containing a poorly constrained small county; read the county’s own score if that county matters. A score of 1.00 beside an estimated value means the published record leaves almost no room for it; it is still an estimate, and the mark says so.
Suppression is heaviest in small counties, which is where most of a county’s values are DTG’s estimates.
Related
The estimation inputs and the constraints are on the methodology’s industry data page; the reliability score is on disclosure and provenance; aggregation is on geography.
Still stuck? A person can help.
Get support