Amy O`Hara, Census - Washington Statistical Society

download report

Transcript Amy O`Hara, Census - Washington Statistical Society

Fully Leverage External Data Sources:
A Census Bureau Change Principle
Amy O’Hara, U.S. Census Bureau
Washington Statistical Society Seminar on
Administrative Records for Best Possible Estimates
September 18, 2014
Vision and Planning to
Effect a Paradigm Change
 Pipedream about making changes in income
estimates in surveys
 Suggestions for moving forward
 Need for vision and long-term planning
 “Requires leadership buy-in at all levels of a
statistical agency”
Change Principle:
Fully Leverage External Data Sources
We will fully leverage administrative records and other
external data sources, including big data, to supplement and
possibly supplant frame development and direct data
collection, enhance self-response capabilities, support new
data products or expand existing data products. Using
external data will help the Census Bureau reduce data
collection costs, redundancies, and respondent burden and
improve data quality. The Census Bureau also has the
opportunity to serve as a data integrator for the federal
statistical system.
- Future of Census Bureau Operations, April 25, 2013
What are we doing to
Fully Leverage External Data?
 Research and testing with administrative
records for use in 2020 Census
 Demographic surveys using administrative
records for frame, contact, modeling, and
 Big Data explorations in Economic Directorate
Strategic Plan:
Research and Implement Records
 Strategic Plan Tactic 1.2.4:
Integrate data from existing sources, including
administrative records, to produce new information
products that provide deeper insights into our people
and economy by combining data sets that had not
previously been linked.
 Strategic Plan Tactic 2.2.4:
Explore and research uses of Big Data.
- U.S. Census Bureau Strategic Plan, FY2013-2017
More from the Strategic Plan
 1.1.3 – Make the 2020 Census more cost effective
while achieving quality targets.
 1.2.7 - Perform world-class research.
 1.6.2 - Use the most cost effective technologies
available to update the MAF and assess its quality.
 2.2.1 - Promote the use of Census Bureau products.
 2.3.1 - Support reimbursable surveys and continue to
improve the Interagency Agreement process.
 3.1.8 - Increase survey and census efficiency by using
empirical data to facilitate intelligent business
decisions prior to and during data collection.
What we are working on
 Integrate new data sources – acquiring state
data, negotiating to obtain new variables,
school frames, housing data
 Big Data – Big Data lab to test concurrent
analysis and estimation, exploring options to
acquire new data sources, Big Data class
Business Plan for Change:
Identify and Acquire Data
 Activity 1.1.2
Introduce new business process for integrating data sets
and producing new data products.
 Activity 2.2.1
Work with the Department of Commerce, the Office of
Management and Budget, and key record suppliers to
establish statistical policies, procedures, and timelines for
maximizing the use of administrative records.
 Activity 2.2.3
Enter into Memoranda of Agreement with key suppliers.
U.S. Census Bureau Business Plan for Change
More from the
Business Plan for Change
 1.1.1 –Identify data gaps and evaluate internal and
external data sets for relevant analyses, data
products, and services.
 2.4.3: Explore options to reduce burden.
 4.3.1 - Establish processes to engage individual and
network partners in Census Bureau knowledge areas.
What we are working on
 Business process to integrate data and create
new products – documentation on data ingestion
and processing, portfolio management
 Policies, processes, and timelines to integrate
data – policies in place, improving
communication and knowledge management to
standardize processes and introduce external
data into operations and products
 Enter MOA with suppliers – new federal, state,
and third party negotiations underway
 Seeking and testing more external data from
federal, state, third party sources.
 Need to:
 Understand data quality
 Design systems to accommodate "big" data
 Collaborate with program and statistical agencies