Improving Government Data Collection

by Jim Hendler and Beth Noveck

Both Congress and the White House have taken initial steps toward creating greater transparency in reporting federal spending.  While preliminary, these efforts could have a far-reaching impact on how governments collect and publish data from the entities they regulate. 

Done right, new rules can create greater transparency and accountability while reducing the paperwork burden on regulated entities. At present, however, both sides' proposals fall short. They fail to recognize that spending is only one type of data collected from players from whom data is repeatedly and inefficiently gathered.

We offer some suggestions for improvement that could lead to reduced compliance and investment costs, improved corporate accountability, greater consumer protection, and will also create new research and reporting test beds to foster data-driven journalism and scholarship about the life of organizations.

The DATA Bill

In mid-June, Rep. Darrell Issa (R-CA) introduced the Digital Accountability and Transparency Act of 2011 (DATA)^[1], calling for quarterly reporting of all federal spending, including grants, contracts and subcontracts by both the recipients and the awarding agencies to an independent successor to the board previously established to oversee the implementation of the Recovery Act. DATA would effectively strip the Office of Management and Budget (OMB) of much of its oversight over federal spending reviews. However, DATA also calls for OMB to set standards about which data elements are to be used in reporting and to follow international, open and non-proprietary models such as XBRL. All reported data is to be published online and, to the greatest extent possible,  the process automated to maximize transparency. 

Executive Order on Accountability

On the same day, the President issued an executive order on Delivering an Efficient, Effective, Accountable Government^[2], calling for agencies to reduce fraud, waste, and abuse and centralizing control over these accountability efforts with the Chief Performance Officer in OMB.  The EO, too, calls for the creation of a new Transparency and Accountability Board, in this case, however, comprising agency personnel from the executive branch.

Suggested Improvements

Without opining here on the jurisdictional debate between branches of government over the control, composition and authority of the new board, there are provisions that could be improved both in the draft legislation and in any subsequent guidance to serve the bi-partisan goals of greater transparency regarding how data is collected and published.

First, the DATA bill provides that items reported include name and address of recipient but no requirement that corporate persons identify the beneficial owner nor any parent-subsidiary corporate relationships. This week the Securities and Exchange Commission proposed draft "know your counter party" rules^[3] for complex financial transactions known as swaps as part of its new package of rules implementing the Dodd-Frank legislation. In the same way, it is entirely doable to add simple provisions to the DATA bill that would mandate disclosure of the ownership and structure of recipients at whatever level of specificity will best enable the public to know who is really receiving the money and how they relate to other recipients.

Second, entities could be mandated to use consistent legal entity identifiers by, for example, picking their corporate entity from a selection list.  This will be useful for building a more consistent, open and standard library of legal entity identifiers within the federal government.  By moving toward a standard list of names in the federal spending domain, we will help agencies to amass a library of common corporate names across different regulatory regimes. Currently, one federal agency might refer to a company as ABC Inc. while another uses ABC Corp. We can help solve this problem by mandating open, universal identifiers here rather than exacerbate it by creating yet another IT system with yet another set of disparate naming conventions.

Third, while mandating a single way of naming a legal entity is important, it's not sufficient to address the fact that every agency also collects different information, ie. names of facilities or securities controlled by that entity. We shouldn't be designing and building a system for reporting spending in a vacuum and focusing only on those limited data elements. Instead, DATA and bills like it should mandate a process that leads toward a single, universal, entity identifier for naming firms with the requirement that additional data fields be open and interoperable.  We want the spending data to be able to "talk" to other data collected about corporate compliance and innovation so we can “mash up” data across agency responsibilities - for example, linking patent activity with the data about federal contracting.  The DATA bill describes only a limited universe of approved standards and the EO is silent on the topic. Instead, any new requirements should mandate the use of non-proprietary, interoperable data elements not subject to any license fees or restrictions on reuse. 

Fourth, data release through the federal data.gov, or via the many data sharing sites being developed by states, cities and tribal governments throughout the US, drives innovation and the development of innovative new startups.  This side benefit to our economy should be augmented in data transparency legislation by allowing thatnew data standards promulgated for use in reporting federal spending should be subject to public consultation, letting developers and others help make sure the systems are open.  The data should also be available in a machine-readable format, to encourage this sharing, with transparency legislation mandating the development of APIs for information sharing.  The DATA bill does recognize the need to allow data to be linked, but it is an ambiguous, throwaway reference to Uniform Resource Identifiers (URIs) in 3612(d)(3)(H). Strengthening this requirement would significantly lower the effort to reporters, economic researchers, and systems developers to reuse this data in our increasingly information-driven economy.

Fifth, there is no authority in either the EO or the DATA bill to create pilot projects and iterate. We don’t understand the problem of inconsistent spending reporting well enough to design -- whether by the legislative or executive branch -- "the" system. Instead, we ought to be allowing small-scale pilots (potentially funded by prize-backed challenges) seeing what works, and trying again.  Further, if the data is made available in machine-readable ways, new systems to make the data more transparent and useful can and will arise outside the government through crowsourced design and use.  This will reduce the development costs while simultaneously allowing more designs to be explored.  In the government, making large-scale “legacy” data systems interoperable is a hard problem that we are trying to retrofit without great expense. This requires more humility and the qualifications to try new policies, technologies, rules and standards. This isn't reflected in the legislation or in the composition and role of the Boards proposed. (We note that the UK’s Data Transparency Board includes a combination of government representatives and outside experts from corporations and academia, and would encourage the US government to consider a similar approach.)

ORGpedia

Outside the government, we have been advocating the creation of an Open Organizational Data Project (http://dotank.nyls.edu/orgpedia), which is committed to assisting with the development of open, interoperable, non-proprietary standards for reporting data collected by government about firms and other corporate entities. With the support of the Alfred P. Sloan Foundation, we are at the beginning stages of thinking through the legal, policy and technology framework for a data exchange that can facilitate efficient comparison of organizational data across regulatory schemes as well as allowing public reuse and annotation of that data.   Currently, we are convening workshops with relevant stakeholders and developing a functional prototype of such a system.  As part of this project, we will also continue to curate feedback on legislative and regulatory approaches to achieving greater transparency, efficiency and accountability.

---

^[1] http://oversight.house.gov/images/stories/Legislation/Leg_Counsel_final_06_10_2011_ISSA_040_xml.pdf

^[2] http://www.whitehouse.gov/the-press-office/2011/06/13/executive-order-delivering-efficient-effective-and-accountable-governmen

^[3] http://sec.gov/rules/proposed/2011/34-64766.pdf

Comments

[Chris Taggart]

Totally agree about "mandating open, universal identifiers" -- these already exist thoughout the world in company numbers issued by company registers, so we (and the governments around the world) already have them -- that's the approach we've taken with http://OpenCorporates.com, and now have over 18 million companies around the world, including 6 US states, with more on the way.

These are in many many ways better than proprietary numbering systems:
1) It's an openly licensed database -- free for reuse
2) We haven't created a new set of IDs/URIs for the company (monopoly ID systems are just plan bad for the community), but ones based solely on the jurisdiction and company number, meaning that if you know the URI you know the jurisdiction and company number and vice versa
3) It's being built from the ground up which the help of the community, following the lead of OpenStreetMap, Wikipedia etc (we may have a long wait for governments around the world to act on this, and make no mistake this is a global issue).
3) Using company numbers is better than the ID methods used by proprietary ID systems as it is a link to the legal entity, which is pretty darned important when you want to sue, fine, or simply get information about it from an authoritative source

And while we'd like the the EPA and other organisations, including in other countries, to use the company numbers/ID systems when referring to them, we already matching federal supplier records, SEC records and IP data, as you suggest.

So, problem solved ;-)

[Tom]

It's great to see you posting about the DATA Act -- and naturally I'm also anxious to see it used as an opportunity to tackle the identifier problem. I'm optimistic on this score: Sunlight has had a chance to speak with Chairman Devaney, and he certainly understands the problem and wants to see it solved. I think you can look at the RATB's paper on award identifiers as a noncontroversial foray into this space -- it doesn't deal with quite the same issue, but it's a strong signal that they're thinking in these terms.

This part surprised me a bit, though:

---snip---
Fifth, there is no authority in either the EO or the DATA bill to create pilot projects and iterate. We don’t understand the problem of inconsistent spending reporting well enough to design -- whether by the legislative or executive branch -- "the" system. Instead, we ought to be allowing small-scale pilots (potentially funded by prize-backed challenges) seeing what works, and trying again. Further, if the data is made available in machine-readable ways, new systems to make the data more transparent and useful can and will arise outside the government through crowsourced design and use. This will reduce the development costs while simultaneously allowing more designs to be explored. In the government, making large-scale “legacy” data systems interoperable is a hard problem that we are trying to retrofit without great expense. This requires more humility and the qualifications to try new policies, technologies, rules and standards. This isn't reflected in the legislation or in the composition and role of the Boards proposed. (We note that the UK’s Data Transparency Board includes a combination of government representatives and outside experts from corporations and academia, and would encourage the US government to consider a similar approach.)
---snip---

While I agree that the composition of the FAST Board could use improvement (there's also the concern over sunsetting), I think the philosophy behind the bill is quite different from what you're describing here. More than anything, DATA represents a move away from prescriptive legislation like FFATA and toward a more agile oversight body that's less subject to political pressure. There are advantages and disadvantages to this approach. The FAST Board would be able to react more quickly and thoughtfully to the realities of the disclosure landscape. But advocates like ourselves could potentially have fewer levers for influencing its behavior. That's worrisome, but on balance I think it's worth a try.

Still, I'm confident that the specific concern you raise -- the ability to iterate and make course corrections -- is very much in line with the thinking of the people behind this bill. It's always risky to put your faith in well-meaning individuals rather than specific statutory language. But prescriptive legislation has proven to be a brittle approach to technology, and I believe in the vision of the folks who'll be staffing the FAST Board. I think the DATA Act is a worthy successor to FFATA.

[Andy]

article that appeared good enough to be read so that adds to knowledge when reading

[Mike]

Is this an update from the The Data Protection Act 1998 commenced on 1 March 2000, with most of its provisions being effective from 24 October 2001. It replaced and broadened the Data Protection Act 1984. The purpose of the Act is to protect the rights and privacy of individuals, and to ensure that data about them are not processed without their knowledge and are processed with their consent wherever possible. subject to more stringent conditions on their processing than other personal data. I feel that LPC
understanding has not been utilized here and the judgment can go against local MP's.

[How Many Calories]

I enjoyed reading all of your posts your blog contain a lot of valuable information. You have done a nice job.

[Calories in Banana]

Totally agree on "procedures that require open, universal identifiers" - they already exist thoughout the world in the company issued by the company records for us (and world governments) have already - this is the approach we made with http:// OpenCorporates.com, and now has over 18 million businesses worldwide, including six U.S. states, with more to come.

These are in many ways much better than the proprietary numbering systems:

1) is a database under a free license - free for reuse

2) We have created a new IDS / URI of the company (a monopoly identification systems are just a bad plan for the Community), but are based solely on competence and the number of companies, that is, if you know you know the URI of the powers the number of the company, and vice versa

3) E 'built from the beginning that will help the Community in the footsteps of OpenStreetMap, Wikipedia, etc. (we may have to wait a long time the government around the world to act on this, and make no mistake, this is a global problem ).

3) The use of numbers in the company is better than the identification methods used in identification systems because it is a unique link to the legal person, which is important when you want to fucking go, well, or simply information for an authoritative source

And even if we wanted the EPA and other organizations, as in other countries, the company uses the numbers / identification systems, when it comes to those who already have a record similar to the Federal seller, SEC filings and IP data, as you suggests.

[company logo design]

The Department of Labor in partnership with the Department of Education announced two billion dollars in grants to support educational and career training programs for workers. Known as the “Trade Adjustment Assistance Community College and Career Training Grant” – or TAA CCCT inside the Beltway – the program is precedent setting in its magnitude of support for 21st century job skill training and for making the default-setting in grantmaking much more open.

[JA]

Much in line with the thinking of the people behind such a bill.

[JIM sare]

It is definitely an open reusable database.Great read.

[nikeshoes8]

Thanks for sharing this nice post.I will keep your article in my idea.

[cheap Polo Shirts]

Hi There is obviously a lot to know about this. I think you made some good points in features also.

[סידור בתים]

This aspect of our economy should benefit from greater transparency of information laws, allowing data that new standards are put in place for reporting on federal spending should be subject to public consultation.

[Maruti Ritz]

Hi, I found this blog once, then lost it. Took me forever to come back and find it. I wanted to see what comments you got. Nice blog by the way.I must take this opportunity to write to you from my heart.I Really enjoyed your blog.I have never read such a wonderful article.

[Canada Goose Parkas]

In my opinion, some government are not responsibility for their people, so they also could not get data well.

[domena]

Great information here, thanks for sharing this valuable information!

[Online Schools in Missouri]

Thanks for the post - capitalizing on game changing technology to increase transparency is key. Investing in the proper technology will allow the government to run more effective practices, increasing productivity for its employees and making it easier for them to share accurate information and data with the public.

[Account Deleted]

Great points mentioned, thanks for sharing this.

http://sevenit.com/Infrastructure-Solutions.php

[cisco ccna training]

I totally agree on how we should make a bill. On our educational days, we were thought many ways if we want to be a public servant but this one, is totally different! It's a friendly tips!

[Oklahoma data center]

Just love it.. its amazing one of the sophisticated post i have read in recent times.

[Online Data Collection]

Hi Thanks for Sharing this post.. I Looking for more post on online data collection. I also have site which provides online service for data collection from multiple sources, securely sharing and online collaboration. Ready online forms & excel like web database to update data plus email notification to act on data.

[iphone apps]

I am hopeful one day there will be no error or if there is, it would be solve right away because of technology and practicing globally about modern ways in data collection.

[Tommy Lilja]

It is a good thing that data collection is been improving in government,data collection and data protection is a very important things that need to be safety in Helsinki Finland there is some data protection services that specializing destroying documents and its for a good thing.

[Bookkeeping Caloundra]

Done right, new rules can create greater transparency and accountability while reducing the paperwork burden on regulated entities.

I strongly agree with your statement. If new rules will be implemented, it will be a great help for government admin. I hope this will be approved as soon as possible.

[beats by dre]

It's really amazing post. I like to read such blog regarding government. Thanks for sharing this one.

[Chicago Bulls Snapback]

The aboriginal Chicago Bulls Snapback mj88jq23 abode Chicago Bulls Hats a lot of MLB Snapback Hats humans attending NBA Snapback Hats and conceivably NFL Snapback Hats the simplest Orlando Magic Hats way to Utah Jazz Hats acquisition the Atlanta Falcons Hats bargain hats Buffalo Bills Hats would be Chicago Bears Hats at the Cleveland Browns Hats administration Dallas Cowboys Hats abundance of Denver Broncos Hats your area.Detroit Lions Hats

[iphone applications]

Generally, this Bill is worth praising! The White House as well as the Congress made a great impact on this far-reaching proposal. Keep it up!