Responding to Federal Data Changes: Connecticut’s Path Forward
While the topic of “disappearing data” may seem alarmist, or perhaps relevant only to data people, what we are learning about changes to the support of our publicly funded data is going to affect every organization, including non-profit and for-profit organizations.
At CTData, we are hosting events to learn together what is happening at the federal and state levels, learn what we can do in our own organizations, and most importantly, identify actions that we can all take as participants in this great democracy.
Last week, over 35 people gathered in Hartford to Bridge Data Gaps and Build Community Connections as part of this work. Some of the highlights from the event include:
Federal data is changing in ways that are not always easy to determine.
Some changes are the result of federal funding cuts to programs, resulting in administrative data not being collected.
Others are the result of staffing cuts that effectively eliminate surveys from being published.
Connecticut is taking steps to protect residents from unconstitutional efforts to access information by strengthening state laws.
Connecticut is strengthening the Shield Law to ensure patient privacy.
Speakers reiterated that it is important to know what is in your state and federal contracts regarding turning over data to state or federal officials.
Much of the data that Connecticut makes available through state agencies are partnerships with the federal government, largely in the form of funding support.
If federal funding is cut, Connecticut may be able to fund some of these programs, but we may lose access to some state data as well.
Connecticut is still moving forward to ensure that race, ethnicity, language, and disability data collection in health settings are aligned with the ways people in Connecticut identify. This work will continue moving forward.
There are many ways that you can take action. The post below includes ideas, and this page includes a list that will be updated as we learn more. And please consider joining us at one of our future events!
What We Learned
What’s Changing at the Federal Level
First, we discussed some of the things we took away from our May 20 event called “Safeguarding Federal Data for Democracy,” which you can access here. Some of the highlights that we shared include:
Trust in sharing data with the federal government is declining as the federal government ignores statute and seeks to access and connect data sources across federal departments.
Loss of federal staff and expertise threatens both the quality and continuity of critical datasets. Data users should approach data sources with increased scrutiny. Data users should also expect to lose access to some data sources, potentially even crucial indicators such as the Federal Poverty Level, as statistical staff are cut.
As federal funding is cut to programs such as Medicaid and SNAP, so, too, will administrative data be lost.
There is uncertainty about how the new federal standards for the collection of race and ethnicity will be rolled out (SPD 15) and uncertainty about their future.
Data is Being Impacted in Four Primary Areas
The state Office of Policy Management (OPM) coordinates with the executive branch on "providing the information and analysis used to formulate public policy for the State," and specifically the Data and Policy Analytics division develops resources the state can use for managing its data. This department coordinates across state agencies. Scott Gaul (Chief Data Officer, OPM) noted that the past few months have shown that data collection is local. States are on their own to figure out what data they need and to ensure they collect that data for themselves.
Data users have observed impact in four primary areas:
Loss of Access to Federal Datasets. States are experiencing disruptions in access to federal datasets, prompting urgent efforts by the private and public sectors to pull, archive, and preserve data (see more in our recent webinar recording). While some resources have come back online, modifications may have occurred without clear communication, leaving uncertainty about data integrity and continuity.
Reliance on Federal Standards. Connecticut’s data collection and analysis heavily depend on federal standards, particularly for demographic and geographic information. A huge amount of effort across the country went into updating the standards of SBD 15. These standards were last updated in 1997, and across the federal government, agencies have started implementing these changes. If federal agencies stop maintaining or updating these standards, it could have major, lasting local impacts and undermine the consistency and comparability of state and local data.
Loss of Federal Staff and Expertise. Significant reductions in federal staff and expertise are less visible and more difficult to find, but deeply impactful. States rely on federal experts for guidance on legal and technical questions. With many federal employees retiring and departments being eliminated, states like Connecticut are left with fewer resources and less support, making it harder to navigate complex data and compliance issues.
Risks to Privacy and Data Sharing. There is a growing risk to individual privacy as federal executive orders seek broader, less restricted access to state-held data, potentially requiring states to share sensitive information about program participants. Particularly troubling are the Executive Orders (EO). One EO calls for information sharing across the government, despite clear guidelines spelled out within agencies and programs on what data should and should not be used for. Sharing data across agencies requires data sharing agreements be in place that outline security and data privacy protections. The other EO removes guardrails on the development of Artificial Intelligence, which relies on being fed large quantities of data. Many organizations are unsure of their legal obligations under these changing rules. It is critical to review contracts and other agreements, consult legal counsel, and understand what data can be shared.
Executive Order 14243, “have full and prompt access to all unclassified agency records, data, software systems, and information technology system” including “to ensure the Federal Government has unfettered access to comprehensive data from all State programs that receive Federal funding.”
Executive Order 14179, “Removing Barriers to American Leadership in Artificial Intelligence” revoked the Biden Executive Order, “Safe, Secure, and Trustworthy Development and Use of Artificial Intelligence,” that sought to encourage Responsible AI development.
State Data: A Critical Resource
As federal data face uncertain futures, it is helpful to understand what state data is available, and what may be affected by federal cuts. Patrick Flaherty from the Connecticut Department of Labor Office of Research and Information gave an overview of some of the myriad data sources the DOL makes available to the public.
Overall, some of the challenges to high-quality data provided by DOL, as well as many other agencies, include:
Declining response rates for surveys, impacting data quality.
Funding shortfalls from federal sources, requiring state subsidies.
Uncertainty about the future of key data sources. For the DOL, that includes key sources such as the Employer Database.
Some of the specific federal-state cooperative data sources Flaherty covered included:
Quarterly Census of Employment and Wages (QCEW): This product is based on Unemployment Insurance (UI) tax filings by employers. It is a federal-state program that relies on both state and federal law and funding, which also means that the data is standard across all states. If the federal funding were to be eliminated, this source, which serves as the basis for other statistics (such as the Current Employment Statistics) would go away unless the state decided to fund the program independently.
Current Employment Statistics (CES): A federal-state cooperative program funded and guided by the Bureau of Labor Statistics (BLS), with some state subsidy due to insufficient federal funding. However, even though the state does subsidize this for Connecticut at the moment, this would be difficult to do at the state. The biggest challenge to this survey is the response rate, which has gotten worse since the COVID-19 pandemic. Additionally, BLS has lost staff, so there are fewer people to follow-up on non-responses. Because of this, the BLS relies more on modeling, and according to Flaherty, when you are using a model, you report on the model, not on what is happening in the economy.
Local Area Unemployment Statistics (LAUS): Another federal-state cooperative program, run in partnership with the Census Bureau and BLS. At the moment, the only potential risk that is known is that staff continues to leave or be cut.
Unemployment Insurance Claims by Demographics & Industry: Data is sourced from UI claims, which is a state-federal program. This source will still be available as long as UI remains.
Data Integration Hub for P20WIN (DatalinkCT) Wage Records: DOL contributes to the P20WIN linking efforts by providing wage records. These rely on UI tax filings and federal workforce program funding; continuity depends on ongoing federal support.
Employer Database: This data source was part of a national contract that has now expired and is not currently not being updated. The future of this source is uncertain.
Some of the sources that have more state funding and support include:
Job Posting Data from Help Wanted OnLine (HWOL): Sourced from a private vendor, not a federal partnership.
Census of Fatal Occupational Injuries: This is intended to be a 50-50 federal-state cost-sharing program, but currently mostly state-funded due to lack of federal support.
Occupational Employment and Wage Statistics (OEWS): A survey of employers conducted as a joint federal-state responsibility. Response to this survey by employers is mandatory, which results in higher-quality data.
Connecticut Occupational Projections: These projections are created through a state-federal funding partnership with the US Department of Labor.
Connecticut Open Data: The state Open Data Portal includes hundreds of datasets. These datasets are compiled according to an ethical standard set by the state so people can’t be re-identified. There are statutory requirements for agencies to provide their data in a secure manner to this portal.
Data Collection and Equity
Sumit Sajnani (Health Information Technology Officer for the state of Connecticut at the Office of Health Strategy (OHS)) affirmed that the state is continuing to improve its support for the ethical collection and usage of race, ethnicity, language, and disability data collected by all state agencies and providers that provide health services. OHS has created resources to ensure providers and agencies are able to collect these standardize data fields, which are consistent with the guidelines developed at the federal level (SPD 15). The Attorney General’s office has reviewed the federal executive orders and has not found any to be impacting the collection of race, ethnicity, and language data collection at the state level.
Sajani shared where these guidelines are now and how OHS is looking to the future:
Connecticut has implemented standardized collection of race, ethnicity, language, and disability data, aligned with federal guidelines, to support equity and health initiatives.
While there has been confusion among providers about whether these standards still apply, Sajnani affirmed that current changes at the federal have not changed the state’s plans to implement these standards.
Due to the increasing mistrust being fostered regarding data collection, Sajnani emphasized the importance of creating a high-quality informed consent process as foundational to building trust for institutional data collection.
OHS is now engaging with providers and the public about how to ethically and effectively use the data that are collected.
What You Can Do: Action Steps for Data Stewards, Providers, and Advocates
Stay Informed and Engaged
Review your organization’s contracts related to state and federal data. Know what you’re required to do if data is requested by federal agencies—consult legal counsel if needed.
Participate convenings hosted by CTData or other events being coordinated by state agencies or other nonprofits.
Strengthen Data Collection Practices
Prioritize informed consent: Clearly communicate to clients and communities how their data will be used, stored, and protected.
Monitor and address declining response rates—engage directly with communities to understand and address their concerns about providing their personal information.
Continue collecting standardized demographic data (race, ethnicity, language, disability) as required by Connecticut law.
Protect Privacy and Build Trust
Create a policy and communicate with staff at all levels of your organization of what to do if a request for data is made. Most staff are not aware of legal requirements of privacy protection and people may target staff members how are least likely to be aware of the law.
Stay up to date on Connecticut’s Shield Law and new privacy protections.
Do not release identified data without understanding the legal requirements and protections in place. Deidentify data whenever possible.
Create an informed consent process if you do not have one, and update your informed consent process if you do. Be sure to include information that has changed based on state and federal law.
Advocate and Report
Report any federal impacts on your programs to CT OPM through their weekly survey and federal impact portal.
Advocate for continued state investment in data infrastructure and workforce to maintain high-quality, reliable datasets. When relevant, respond to federal registers regarding data collection and security. Information on these can be found at some of the resources provided on this page.
Share this information
If this information has been valuable to you, please consider sharing with a colleague who may find this relevant.
Looking Ahead: Join Us at a Future Event
If you’re disappointed that you missed this event, you still may be able to join us at a future event! Sign up for our newsletter or bookmark this page to get alerted when new events are posted. We will be holding events in New Haven, Norwalk, New London, and Waterbury over the course of this year. And save the date of December 3 for our annual conference, which will include more information on this topic.
For More Information
Head on over to our Data for Democracy page for resources and other events coming your way. Save the date for our annual conference, which will explore these topics further. You can stay up-to-date on the latest data and tools by subscribing to our newsletter and following CTData on Linkedin, Instagram, and Bluesky.