Plan - Build a strategy, governance structure, and inventory of data.
Establish a plan
Example 1: Connecticut:
Not later than December 31, 2018, and every two years thereafter, the Chief Data Officer, in consultation with the agency data officers and executive branch agency heads, shall create a state data plan. The state data plan shall (1) establish management and data analysis standards across all executive branch agencies, (2) include specific, achievable goals within the two years following adoption of such plan, as well as longer term goals, (3) make recommendations to enhance standardization and integration of data systems and data management practices across all executive branch agencies, (4) provide a timeline for a review of any state or federal legal concerns or other obstacles to the internal sharing of data among agencies, including security and privacy concerns, and (5) set goals for improving the online repository established pursuant to subsection (i) of this section. Each state data plan shall provide for a procedure for each agency head to report to the Chief Data Officer regarding the agency’s progress toward achieving the plan’s goals. Such plan may make recommendations concerning data management for the legislative or judicial branch agencies, but such recommendations shall not be binding on such agencies.
(f) Information technology-related actions and initiatives of all executive branch agencies, including, but not limited to, the acquisition of hardware and software and the development of software, shall be consistent with the final state data plan.
While Oregon’s CDO is required to establish an “enterprise data and information strategy,” Connecticut is the only state required by law to have a formal state data plan. The plan was developed in a series of iterative phases, with each providing an opportunity for input from the public and government agencies. The plan consists of a series of strategic principles and practices to guide agencies in the use and management of data, along with public policy issues called “Focal Points” to establish priority areas for state agencies to focus data efforts.
Establishing a data governance structure - No state currently has a legislated data governance structure however the Commonwealth of Virginia has established one by executive order.
Inventory Data -
Example 1: Florida
The Chief Data Officer, in consultation with state agencies shall develop an enterprise data inventory that describes the data created or collected by a state agency, including geospatial data used in a state agency’s geographic information system, and recommend options and associated costs for developing and maintaining an open data catalog that is machine-readable. For purposes of developing the inventory, the Chief Data Officer shall establish a process and a reporting format for state agencies to provide an inventory that describes all current datasets aggregated or stored by the state agency. The inventory shall include, but is not limited to: 1) the title and description of the information contained within the dataset; 2) a description of how the data is maintained, including standards or terminologies used to structure the data; 3) any existing or planned application programming interface used to publish the data; 4) a description of the data contained in any such existing interface; and 5) a description of the data expected to be contained in any currently planned interface.
Example 2: Connecticut
On or before December 31, 2018, and not less than annually thereafter, each executive branch agency shall conduct an inventory of any high value data that is collected or possessed by the agency. Such inventory shall be in a form prescribed by the Chief Data Officer. In conducting such inventory, data shall be presumed to be public data unless otherwise classified by federal or state law or regulation. On or before December 31, 2018, and not less than annually thereafter, each executive branch agency shall submit such inventory to the Chief Data Officer and the Connecticut Data Analysis Technology Advisory Board.
Both examples provide a relatively straightforward approach to establishing data inventories. Florida’s law includes some details regarding the minimum amount of information should be collected as part of the inventory while Connecticut’s provide fewer guidelines. Additionally, Connecticut’s law requires that the inventory be updated on an annual basis while Florida’s contains no such requirement. Connecticut’s law is also limited to inventorying “high value data” which means certain datasets such as spreadsheets maintained by an individual for non critical purposes may not be captured.
Both Florida and Connecticut used a similar template to collect their data inventories and both were able to complete the process within a few months. The inventories collect high level information about various state data assets that range from large systems like Medicaid eligibility and claims, to geospatial datasets. Connecticut’s inventory is publicly available and has been updated for 2019