Data Inventory and High-Value Data Release Plan

PDF version of this report

Introduction

Social Security celebrated its 75th birthday on August 14, 2010. Over the last 75 years, we have collected data to carry out our mission. Our data are about people-their wages, their identifying information, their employers, their addresses, and much more. The first regulation we published included a commitment to the public to safeguard the personal information entrusted to us. This commitment is as solid as it was 75 years ago and is further strengthened by privacy laws. We cannot publicly release much of our data because it is protected by the Privacy Act, the Internal Revenue code, the HIPAA, and other statutes. While some of the data can be anonymized, much of it cannot. Our inventory planning recognizes these constraints and all releases will protect privacy in accordance with all applicable laws.

New Open Government Requirements

In the first days of his administration, President Obama issued a memorandum for the heads of executive departments and agencies announcing a commitment to transparency and open government. In this document, the President instructed the Director of the Office of Management and Budget (OMB) to issue instructions for implementing the open government principles. The Open Government Directive (M-10-06) released on December 8, 2009 provided specific requirements agencies and departments must follow in order to fully adopt the open government principles. It also mandated the development and publication of an Open Government Strategic Plan.

Under the transparency principle, OMB directed agencies to explain in detail in their Open Government Plan how they would improve transparency. Specifically, the Directive said, "To increase accountability, promote informed participation by the public, and create economic opportunity, each agency shall take prompt steps to expand access to information by making it available online in open formats." We recognize that foundational steps to becoming more transparent and increasing accountability is to inventory agency high value information currently available for download and to identify high value information not yet available and establish a reasonable timeline for publication online in open formats¹.

Actions Taken

We have completed a number of actions to increase transparency by making information available at ssa.gov and Data.gov.

• In response to the launch of Data.gov in fiscal year (FY) 2009, we began our inventory of data and released two datasets in May 2009.(See last two items in Chart 1.) These datasets met government standards for Data.gov as well as our privacy and disclosure review processes.
• During 2009, we continued our inventory of data. We addressed disclosure and privacy issues, in part, by establishing a point of contact in the Office of the Chief Information Officer for review of each dataset prior to release to Data.gov. We established an executive steering committee for oversight of all open government activities.
• In January 2010, as a result of our ongoing inventory process, we released additional, high value datasets in accordance with the Open Government Directive.
• On February 6, 2010, we launched our new Open Government Webpage (ssa.gov/open), containing datasets and information about our overall management and organizational structure. Through this webpage, we also launched the IdeaScale tool that we used to obtain public input on our planned transparency and data initiatives. After a period of public engagement (that included the IdeaScale tool feedback, advocate meetings, and researchers input), our executive steering committee evaluated the input and recommended goals, initiatives, objectives, and milestones.
• On June 24, 2010, we published our Open Government Plan which included a list of datasets already posted or planned through FY 2011. In this document, Chart 1 contains datasets already posted and Chart 2 lists planned dataset releases through 2011.
• The Open Government Plan includes five major objectives that support the goal of increasing transparency. One of the critical activities is a data inventory, which was completed in September 2010, the milestone shown for this activity in the Open Government Plan.
• In response to the Open Government Directive, we named an Executive Accountable for Publicly Disseminated Federal Spending Information Integrity. We also released our Data Quality Plan for Federal Spending Information on May 14, 2010. The Data Quality Plan outlines our strategy beginning with the certification of our financial data, extending to our other data as well. The long-term strategy consists of a data quality framework that includes:
• governance structures (e.g., disclosure review boards, executive steering committee, working groups, etc.) and processes;
• risk assessments;
• control activities;
• communications; and
• monitoring.
• We established a working group of subject matter experts that meets to review release timeframes, quality, security, and privacy. Following the leadership of the Program Management Office of Data.gov at the General Services Administration (GSA), we developed a dataset checklist sheet modeled after the one GSA uses (see Appendix). Included in the checklist is the foundational element of quality as by the Information Quality Act: https://www.fws.gov/informationquality/section515.html.
• We issued implementation instructions that address data utility, objectivity, integrity, transparency and reproducibility, and made this publicly available at: www.ssa.gov/515/ssaguidelines.html
• We examined our disclosure review board process to ensure it conforms to Data.gov standards. We connected the review board management and our Chief Privacy Officer with the OMB and White House Working Group on Privacy and Security. They ensure that we are aware of national security and privacy issues in the Federal community, enabling us to take appropriate action.
• We refreshed our understanding of the legal limitations of sharing certain data such as master earnings file data protected under the Internal Revenue Code.

List of Datasets in Open Government Plan

We posted 22 datasets on Data.gov. All datasets relate to our core mission and align with our strategic goals and performance measures. The following information in Charts 1 and 2 shows the datasets already posted as well as those scheduled for release by the end of FY 2011.

Background on Enterprise Data and Our Business Intelligence Architecture that Impacts Data for Data.gov

As our program responsibilities grow, data and databases also grow and change. Historically, we developed our programmatic software at different times for different purposes, and deployed it in a stovepipe environment². In order to give a comprehensive and transparent picture of our data, we will need to derive datasets for Data.gov from several sources (e.g. Case Processing Management System, Appeals Review Processing System, etc.) as we continue conversion from our older, homegrown database structures {e.g. Master Data Access Method (MADAM)} to modern relational database technologies.

Going forward we will explore incorporating data transparency considerations into the systems development lifecycle so that providing high value data in standard formats for public use is a more efficient, automated process. Building steps into the lifecycle involves working with and through several governance groups. For example, the lifecycle Change Control Boards determine the appropriate placement of each step in the lifecycle relative to other tasks and data needs; the Deputy Commissioner for Systems Management Steering Committee determines the resource and schedule impact on project activities. We tentatively plan to identify a selection of pilot projects and evaluate their results prior to implementing any significant lifecycle changes.

In addition to the database strategy referenced in the paragraph above, we have made progress in overhauling information sharing through continued development of a business intelligence (BI) architecture. The process began in the late 1990's with internal collaboration and the use of technology solutions to address data integration and workload management challenges. Once the BI architecture was in place, we transitioned workloads incrementally, workload by workload, based upon value to us, and governed by our architectural review board. Master data management is an important component of building the BI architecture and provides consistency and data quality³. It is through the BI architecture that we will be able to produce summarized datasets for Data.gov.

Methodology for Obtaining Datasets for Data.gov

Chart 3 shows three different methods for deriving datasets for Data.gov. Line 1 reflects the use of the BI architecture. Using master data management in the BI architecture, we summarize and aggregate data, making it available for Data.gov without PII. Besides the datasets available through the BI architecture model, we produce additional datasets through data extraction from the program specific data we maintain. Line two shows how we use anonymization methods to develop public use files as well as summarized reports. Line three illustrates a similar process to produce public reports from the survey data we collect, along with program specific data we maintain.

While not reflected in the Chart 3 diagram, we also produce datasets in response to Freedom of Information Act (FOIA) requests. Since we generate the data and information in response to individual requests, there is not just one method for responding to FOIA requests. We look at each information release to determine if it is appropriate for Data.gov.

Chart 3 - Examples of Open Government High Value Dataset Development⁴

Additional Datasets for Data.gov Based Upon Inventory of Existing High Value Data

Based upon a review of our data, we identified high value data categories, shown below, that will improve public knowledge of our operations and our programs. For all of the categories, we will extract reports that contain summarized data so that even if the public combines our datasets with other data, no violations of privacy can occur. We will verify and certify the quality of the data in accordance with the Information Quality Act and the Data Quality Framework developed by our Executive Accountable for Publicly Disseminated Federal Spending Information Integrity. In addition, our Open Government Executive Steering Committee will advise, and as needed, concur, on the appropriate level of data (local, regional, national, etc.) and frequency of release (e.g. weekly, annually, etc). As needed, we will refer datasets to disclosure review boards for clearance should there be any questions regarding security and/or privacy. The Checklist for Data.gov Submission (see Appendix) will be signed off by the appropriate officials before the data are made available to Data.gov.

The following list is grouped according to major mission and program areas, not by priority. These areas have high value data that are available and not released. These areas supplement the extensive data already released on Disability Determination Services and Office of Disability Adjudication and Review case processing, as well as statistical information that our researchers make available on a regular basis. Our first priority for earliest release, in FY 2012, will be the data that have already been determined to be the "official agency measure" in our Unified Measurement System and available through the BI Architecture, as well as data in support of Performance.gov metrics. The second priority will be budget information, in consultation with the Senior Accountable Official on the Quality of Federal Spending Information. The third priority will be administrative information that includes data about staffing and human resources. By focusing efforts in these areas, by the end of FY 2012, we will provide a transparent picture of program activities, our budget, and our staff. We will also release as soon as possible, without respect to the above priorities, those datasets requested by our key audiences. (See our Communications Plan, Appendix C, in the Open Government Plan.)

The following chart reflects the inventory of high value data not released, but available as explained in the paragraph above.

Chart 4--Future New High Value Datasets
(Release Begins in 2012)

Service

• Field office visitors
• Field office claims appointments scheduled within 21 days
• Public satisfaction with field office and 800 number service

Initial Claims

• Field office claims pending
• Number and percentage of disability claims filed via the Internet
• Disability Determination Services production per workyear
• Disability Determination Services case processing time with health information technology
• Old-Age, Survivors, and Disability Insurance and Supplemental Security Income case accuracy and dollar accuracy
• Disability Determination Services net accuracy

Hearings

• Appeals Council and Court remands sent to hearing level
• Hearing production per workyear total and decision writing
• Hearings pending by electronic or paper folders

Appeals Council

• Appeals Council request for review receipts, dispositions, pending
• Appeals Council production per workyear
• Court remands sent to Appeals Council

Court Cases

• Court level receipts, completed and pending

Earnings

• Annual earnings items completed by electronic and paper
• Forms W-2 completed (all formats)

Other Workloads

• Continuing disability review case outcomes, completed and pending
• Supplemental Security Income redeterminations processed compared to budgeted target for the year
• Enumerations completed, processing time, and enumeration at birth completed
• Medicare Part B and Part D workloads
• Claimant representative data
• Old-Age, Survivors, and Disability Insurance notice data
• Limited English Proficiency data

Administrative

• Social Security staff on duty with impact to national and state economy
• Disability Determination Services staff on duty with impact to national and state economy

Other

• Performance.gov metrics data
• Administrative cost data by appropriation and allotment, which may include the following (Disability Determination Services budget data already provided for in Open Government Plan Appendix A):

Information Technology Systems	Social Security Advisory Board
Reimbursable activity	Automation Investment Fund
Delegated Buildings	Disaster Relief
Construction	Recovery Act funds
Low Income Subsidy	Office of the Inspector General
Continuing disability review allotments	Other allotments and targeted appropriations
Supplemental Security Income non-disability redetermination allotments

Additional SSA Data-Fostering the Use of Data and Other Traditional Releases

1. The Research and Statistical Community

Continuing our long tradition of work in the research community, we will develop and release additional statistical tables, studies and research working papers as part of our ongoing mission. Research and statistical products by our Office of Research, Evaluation, and Statistics are available at: Policies. Our regular publications include, for example, the Annual Statistical Supplement to the Social Security Bulletin. The Supplement is a major resource for data on the nation's social insurance and welfare programs. The majority of the statistical tables present information about the programs we administer. In 2009, our Office of Research, Evaluation and Statistics began a 3-year initiative to modernize how we produce statistical tables and publications. The new process will include the development of summary datasets that can be converted to machine readable format (e.g., comma separated value). Consistent with our commitment to transparency, we will include the new datasets in the Data.gov inventory. As reflected in our Open Government Plan, as well as Chart 2 of this document, we plan to release the first dataset based upon this new modernized process in FY 2011. As the modernization continues, and new datasets become available, we will review and include them in the Data.gov inventory.

One of our ongoing initiatives that foster the use of our data is the Retirement Research Consortium (RRC). The Retirement Research Consortium (RRC) consists of three multidisciplinary centers housed in three separate institutions (Boston College, the University of Michigan, and the National Bureau of Economic Research). We fund the activity through cooperative agreements and awarded approximately $7.5 million to the RRC in FY 2009, the last year of the current 5-year award. We expect funding to continue at that level for each of the remaining years of the award.

The RRC has three main goals:

• Conduct research and evaluation on a wide array of topics related to Social Security and retirement policy;
• Disseminate information on Social Security and retirement issues relevant to policy makers, researchers, and the general public; and
• Train scholars and practitioners in research areas relevant to Social Security and retirement issues.

To meet these goals, the centers perform many activities. They conduct research, prepare policy briefs and working papers, hold an annual conference, and provide research and training support for young scholars. Recent RRC research is provided via this link: http://www.ssa.gov/policy/rrc/subjects.html

2. Other Traditional Releases from the Office of the Chief Actuary

The Trustees Report and other actuarial information is available at: http://www.ssa.gov/OACT/pubs.html

Other useful data is currently available from the Office of the Chief Actuary using the following link (the link provides drill down capability (for 1967 on) for information on the number of beneficiaries, average benefits, and type of beneficiaries): http://www.ssa.gov/OACT/ProgData/icp.html

Chart 5- Information That May Be High Value but Not Currently Available
(Planning for Release in 2013 and Continuing)

• Data from public engagements and surveys and other social media
• Citizen Authentication Statistics
• Facility Information (subject to national security considerations)
• Geographic Information System Data Presentations
• Information Technology Hardware and Software Inventories
• Representative Payee
• Debt Management

Concluding Thoughts

This document provides the results of our complete data inventory in accordance with the Open Government Directive. However, our information and data will continue to evolve. Therefore, we will need to refresh the inventory periodically as new programs, applications, research, surveys, and public engagements change our data holdings. In some cases, we will retire and consolidate data holdings.

In addition, until the launch of Data.gov, we developed our methods and processes for gathering data, extracting data, and displaying data to support internal management, operational, and executive needs. For example, within the organization today, we do not build data gathering for public release into the project life cycle, and extraction methods have been primarily for executive decision-making or specific research. Today, where the data are available, we will reformat and present it in ways that benefit the public, while at the same time protecting their personal information.

These methods will need to evolve to incorporate transparency of the data (i.e. public release in standard formats) into the requirements for new data gathering efforts, with an ultimate goal of operating with a more agile process that considers both internal and external use of data in the future. In the future, we will specify the format and presentation of the data upfront so that preparation of data for public release will be both quicker and easier. We are exploring putting the appropriate tasks into the Planning and Analysis (P and A) and Construction phases of our systems project life cycle to identify and verify the data that would be published as well as to conduct data validation and certification. Our goal is to integrate this process with other appropriate data, testing and certification activities rather than add another stand-alone activity.

Appendix - Checklist for Data.gov Submission

NOTE: This checklist is the culmination of data review and oversight processes and is completed by the data sponsor with sign off by designated officials. Additional information for any item may be included in a supplemental page and attached to the form.

End Notes

1 High value information is information that can be used to increase agency accountability and responsiveness; improve public knowledge of the agency and its operations; further the core mission of the agency; create economic opportunity; or respond to need and demand as identified through public consultation.

2 Social Security Administration, Information Technology Vision 2009-2014, February 2009, page 14.

3 Case Study: BI Strategy Prepares US Social Security Administration for the Future, Gartner Research, Bill Gassman, July 31, 2009, page 4.

4 This chart does not include datasets produced in response to FOIA requests. The numerous datasets related to our budget process, Annual Performance Plan, Performance.Gov and Priority Goals are also not included.