By Dennis D. McDonald

What I found most interesting and useful about the Data Transparency 2013 conference held September 10 in Washington DC was the diversity. The ability of Hudson Hollister’s Data Transparency Coalition to bring together such a high-quality group of speakers and attendees representing so many different perspectives was quite an accomplishment.

I’m still digesting my conversations and  the notes. What follows are, for me,  high points and impressions. I’ll be writing and talking about individual points in more detail in the coming days.

Conference agenda

From the conference announcement:

• Data Transparency 2013, hosted by the Data Transparency Coalition, will be the nation’s first federal open data policy conference. For the first time, the leading advocates of these policies - from the executive branch, the legislative branch, the nonprofit sector, the tech industry, and the financial industry - will gather to compare the history and chart the future of open data.

My observations

1. Today much movement exists towards more open access to data generated by federal programs. Pressure is coming from both the legislative side and the executive side of government and from several different public and business oriented interest groups.
2. The importance of standardization is universally acknowledged as a facilitator both for interoperability as well as analytics.
3. Practical issues related to governance, costs, training, and privacy/sensitivity are impacting progress in making data accessible, even when open access is demanded by legislation or by OMB.
4. The central role of agency IT departments in managing data access must be acknowledged. At the same time there is a need to understand how open data programs support agency and program goals and objectives.
5. Without more centralized control and leadership, the variety of federal agency programs that are being developed at OMB direction could actually slow the move toward more standardized approaches to making federal program data assets more publicly available; see Project Open Data’s Implementation Guide and the various tools provided by GSA and OMB that support inventory development, metadata capture, API development, and catalog development.
6. Justifying public access to program data in hopes that valuable new products and services will evolve will eventually forced a review of cost versus benefits of open data programs to the taxpayer. They will also eventually force a formalization of the definition of public versus private sector roles in carrying out government programs.
7. We have not yet completely understood and internalized the lessons of the Recovery Accountability Transparency Board. Despite — or perhaps because of — its “crash” nature it still has much to teach us about making public data accessible.
8. If more standardization and control are not exerted over how current data sets are being  inventoried and described by Federal agencies, many of the potential advantages of open data could be at best delayed or at worst lost. (The manner in which the Open Data inventories are developed between now and November will be one indicator of the degree to which such efforts can be decentralized.)
9. Agencies like Interior that historically have been responsible for physical objects and events have a significant advantage in terms of standards, governance, and technology in making their data publicly accessible. One question is, how generalizable are the lessons we can learn from them?
10. There appears to be an inherent conflict in the need of independent agencies for specialized data management infrastructures and the potentially competing desire to provide universal access to data across agency program. Figuring this out may require significant data modeling and data definition efforts. Can we afford the wait?
11. The shift from report-driven financial reporting to data-oriented financial reporting will require much exercise of power and influence by a central financial authority.
12. “Intelligent data” and “open data” are not the same thing. Are there  inherent conflicts between the goals of (1) adding structure, meaning, and context to data and (2) making data accessible to the public at the same time?
13. We need to get the numbering of companies and organizations under control. This has to be an international effort. Is the federal government the best organization to drive standardization and US participation?
14. Legislative requirements for open data need to acknowledge and accommodate governance and cost implications at the agency and program level. Information systems, especially those designed to engage with the public, require time, money, and expertise.

Discussion

In addition to the diversity of interests — which is good —  I’m seeing real progress in making government data more accessible. That some legislative efforts have bipartisan support is not a trivial concern. As an information system professional, though, I still think there are several warning signs we need to pay attention to:

1. Imposing standards and building new systems costs money. Are the Congress and the Executive realistic about this? Given the ongoing impacts of sequestration I have my doubts. I sincerely hope that the efforts of Executive agencies to inventory data sets and develop policies and strategies for engaging with the public about these data do not, because of such issues, get sidetracked.
2. The current Administration actively supports open data programs. Will this support continue when the Administration changes?
3. There seem to be at least two visions at work that promote open data. One vision is that data will be made available to third parties who will be able to do new and innovative — and valuable — things with these data. Federally supplied weather data and other geocoded data sets are regularly pointed to as examples of how new industries and jobs can spring up around data that are “freed up.” Another vision is that the role of data within agency programs will become second nature and integral to the accomplishments of agency goals and objectives. Are these two visions compatible? I’m not sure and I’ve tried to address this issue at a strategic level in A Framework for Transparency Program Planning and Assessment.
4. We’re dealing with some fiercely independent bureaucracies within the Federal Government. Can we overcome the inherent insularity and the real costs of addressing the many legacy systems that are in the process of being re-architected?
5. Most important of all, how supportive is the public of open data programs? And will the public be willing to pay the price?

When all is said and done, though, I am very optimistic!

Copyright (c) 2013 by Dennis D. McDonald