Mar 6

Mar 6 USAID’s “Frequently Asked Questions” and the Management of Open Data Programs

Agency Data, Data, Data Access, Development, Governance, Open Data, PMO, Process, Project Management, USAID

I wrote about The U.S. Agency for International Development’s open data program last December in “USAID’s Evolving Open Data Culture” (/usaid.html). USAID recently published a FAQ document that describes the workings of its open data program regarding data generated by funded programs; see the USAID’s “Frequently Asked Questions” (http://www.usaid.gov/data/frequently-asked-questions ). Among the topics covered by the FAQ are:

The kinds of data that must be submitted
Requirements for machine readable datasets in documented nonproprietary formats
Ownership
Partner relationships
Which development contracts are affected
Treatment of sensitive data
The role of “data stewards”
Description of the data submission process
Conditions under which submitted data might not be released
How data will be made available

Not a lot of detail is provided but several links are included to additional and more detailed information including formal statements of contract requirements.

One of the most interesting related documents is “ADS Chapter 579 USAID Development Data” (http://www.usaid.gov/sites/default/files/documents/1868/579.pdf) which provides more detail about several of the USAID open data program’s most interesting components including:

“Dataset access Levels” are described as Public, Restricted Public, and Non-Public (579.3.2.4 )
Roles and responsibilities in the Data Publication Process (579.3.2.5 )
A fairly straightforward definition of “machine readable” (579.3.2.6 b)
A description of applicable “embargoes” if the Dataset “… is the subject of a pending publication or pending patent application.” (579.3.3.3)

This being a government document there are many references to rules and regulations having to do with legalities and acquisition regulations.

What I found most useful from a program management perspective were descriptions of roles and responsibilities as submitted datasets make their way through the overall journey from submission to being made public. These make you realize that an open data program has to interact with a variety of different operating entities both internal and external to the sponsoring agency.

How this overall process is managed throughout the entire “open data program management lifecycle” will determine how effective the program is, how much it costs to operate, and how long it takes.

Something I am currently researching is how to best govern an open data program where a complex mix of technical, business, and legal resources, drawn from a variety of organizations, need to be coordinated.

By itself, just documenting a process is never enough. Also needed is a sufficiently-resourced governance structure that supports program management in how systems and processes are changed and in how they operate.

Related reading: