Skip to main content
Preserving Information in Government

Alan Kowlowitz, Manager, Electronic Records Management Services NYS Archives and Records Administration

Alan Kowlowitz gave a presentation on data preservation in government agencies. He started by listing a series of problems in data preservation:
  • Preserving access and usability over time requires ongoing maintenance of the data, which can be costly.
  • The costs are acceptable for information with immediate business value.
  • The issue of preservation is usually not addressed when systems are planned.
  • The information without immediate business value is at risk.
He then gave a few examples of information resources lost due to lack of attention to long term preservation issues. He presented two examples at the federal level: the 1960 Census and the National Aeronautical and Space Agency (NASA) information that had been stored on equipment no longer available. At the state level, in New York, he gave the examples of the Land Use National Resource System (LUNRS), the Children Youth Management Information System (CYIMIS), and the Committee on Sentencing Guidelines. Information generated by all three programs is no longer usable.

He presented a series of reasons for data preservation:

(1) a clear legal requirement, it is more easy if the agency has a mandate;

(2) a long-term programmatic need, i.e. if the agency see a long-term use;

(3) existence of secondary users, it increases the value of information and the risk of managed information;

(4) high risk and visibility, agencies want to be able to answer questions;

(5) demonstrable benefits; and

(6) organizational culture.

Alan gave a number of examples of organizations that have successfully addressed their data preservation issues: DOH Vital Records/Disease Registries, DCJS Trends Data, DED BEDS, and DOCS Under-Custody. The future direction for SARA is to address data preservation issues, specifically how to reduce costs through the use of more focused system planning, technology, organizational models, and education.