Establishment statistics

Establishment statistics is the area of statistics where data are either about or for establishments. Establishment in this context is a general term that encompass entities other than individuals and households, which characterise social statistics. In establishment statistics, entities are typically businesses, organisations, institutions, and similar. Establishment statistics is related to business statistics and economic statistics, but is wider than these as the range of unit types dealt with in the former is wider. Because of specifics of these units and methods needed to address the issues, establishment statistics has grown into an own field of statistics.


(based on sampling theory, grown out of (official) statistics production as developed since 19nn, ... ...)


<this what follows is an attempt to transfer a figure into a structure here - perhaps pushing too much a specific view (BL)>

Core areas: sampling

survey strategy (choice of sampling and estimator)

sample size considerations


coordinated sampling

multiple frames


Small Area Estimation

nonresponse treatment

Wider approaches

Design of statistical programs

optimisation over several study variables

metadata driven design

Data collection

Questionnaire construction and evaluation

Understanding and enhancing motivation for participation in business surveys

Response process in business surveys

New data sources

  • xbrl
  • businesses’ admin systems
  • unstructured data

Measurement process and measurement error

  • interviewer effects
  • mode effects
  • mixed modes

Business Register creation and maintenance

  • coding
  • system of units & unit relations
  • profiling
  • the unit error/unit problem


  • feedback from editing to design
  • selective editing
  • automated editing
  • probabilistic editing


  • Nowcasting/flash estimates
  • Network approach(es)

Dissemination of statistics

  • Geographic Info Systems


Approaches from a total survey error perspective

Relations between error sources

Theory for admin data integration

  • two-phase life-cycle model

Adaptive data collection (in e.g. IBSP)

Systems for business and economic statistics production

  • ABS
  • Statcs Can (IBSP)
  • Stats New Zealand

TQM (costs, quality, risk)


Meeting user needs/New subject matter statistics areas


  • global value chains
  • international trade statistics

Factoryless production

Intellectual property


  • No labels