Innovations in Data Dissemination Thomas L. Mesenbourg, Jr. Acting Director U.S. Census Bureau


Innovations in Data Dissemination

Thomas L. Mesenbourg, Jr.

Acting Director

U.S. Census Bureau

United Nations Seminar on

Innovations in Official Statistics

February 20, 2009

Dissemination Challenges

Present relevant and appropriate information that meets users’ needs

Provide information that is timely, accurate, and accessible


Topics to be Discussed

Integrating data from multiple sources

Allowing users to access, manipulate, and generate results themselves

Points for discussion


Integrating Data from

Multiple Sources

TheDataWeb – U.S. Census Bureau

– Open source system – handles different kinds of data and datasets in the same system framework

– Empowered by statistical intelligence

• Documentation, statistical usage rules, data integration rules

– Stores data one time, uses them many times

– Software infrastructure that makes HotReports and DataFerrett work

– Collaborating with ABS since 2005


U.S. Census Bureau’s HotReports

Target decision makers who have limited time and statistical background


– Bring together relevant variables for local areas

• Maps, charts, simple tables, and short text boxes

– Focus on one topic using information from many data sources

• Such as economic development or emergency response

– Update the information dynamically

– Support decision making through guided use of statistical data



Local Employment Dynamics Program

U.S.Census Bureau

A collaboration between the U.S. Census

Bureau and almost all the states

– Integrates workers’ characteristics with their employers’ characteristics

– Labor force information available quarterly and longitudinally for small geographic areas and by industry

– Pre-defined and user-defined tables

OnTheMap software

– Easy visualization of massive amount of data

– Strict protection of respondents’ confidentiality

– First released in 2006 – latest version 3.2 released in December 2008




Integrating Data – Other Countries

Two innovations (Sweden)

– Aggregated statistics – SSD

• Access data from Statistics Sweden and other agencies

• Combine data with different structures on same topic

• Transmit selected tables to the user

– The Planning Portal – integrates geographic information from various sources for planning purposes

Interactive databases (Brazil)

– Dynamic graphs that allow adding a time variable

Layered approach (Australia)

– Headlines, simple stories, detailed stories, data cubes, and related data, as well as metadata


Allowing users to access, manipulate, and generate results themselves

Public-use products, including microdata

Statistical data enclaves

User-defined tables while protecting confidentiality

Licensed access to confidential data


Public-Use Microdata

DataFerrett is an analytical tool developed by the U.S. Census Bureau


– Targets sophisticated analysts who are now able to access many public-use microdata files at once

– Data manipulation, including advanced tabulations

– Maps and business graphics using statistical rules

– Data extracts in popular statistical package formats

– Adding regression and other advanced statistical tools

Provides context with the variables

– The universe, variable values, and the definition


More on Public-Use Products

Anonymised micro data files (Ireland)

– No identifiers, and data suppressed if reidentification would be possible

Output Database (Lithuania)

– A specialized web site of Eurostat products

• Developed for national users, free of charge

• Launched with 820 datasets


Statistical Data Enclaves

The U.S. Census Bureau

– Research Data Centers allow users to obtain special sworn status to access confidential data in controlled locations for approved purposes

Also used in Canada, New Zealand and the



User-Defined Tables and Analyses

While Protecting Confidentiality

On-line statistical database (Ireland)

– Users can create customized tables

– Access to software for downloads

Microdata Analysis System (US)

– A remote access system to create tables and execute regressions using confidential data is in development

– Based on the Advanced Query System developed for Census 2000 data


Licensed Access to Confidential Data

Microdata on line (Sweden)

– Approved researchers access confidential data on line

Licensed access to confidential data (Australia)

– Restricted information when access is broader, and more detailed variables only for trusted users

Research Microdata Files (Ireland)

– Only authorized people can access them, under the Statistics Act of 1993


Points for Discussion

If we build it, will they come?

Are statistical agencies best equipped to satisfy users needs?

Targeted versus enterprise-wide systems?

Can innovative, targeted systems be adapted for the enterprise?

How can we leverage each other’s initiatives?