Neglected transparency?

[TPP Weil]

Archon Fung, Mary Graham and I co-direct the Transparency Policy
Project at Harvard's Kennedy School. , Mary published this article
last Saturday in the Boston Globe regarding some of the neglected
transparency issues facing the administration.

http://www.boston.com/bostonglobe/editorial_opinion/oped/articles/2009/04/25/transparency_and_the_right_to_know/?page=full


We would be interested in the groups reaction to the op-ed and our
take on another part of the transparency agenda we argue requires
attention. Specifically:

We make a distinction between transparency policies created to enhance
the public's right to know about their government from specific
policies that seek to achieve a particular goal. Although some
challenges cut across both types of transparency (accurate information
conveyed quickly and in useable formats), others are quite
distinctive. For example, transparency to reduce social risks require
close attention to how users of that information think about and
process that information and how, in turn, that information fits into
actual decisions. Do we have these distinctions right? How engaged
is the right to know community in issues related to more targeted
forms of transparency like those Mary discusses?

Are there discussions of some of these focused transparency policies
going on within the administration / agencies?

Are other groups pushing in similar directions to expand the
transparency agenda to include improving these policies?

Looking forward to your thoughts and reactions.

[John Wonderlich]

Welcome!

I'm trying to process the distinction from the op-ed.

I see more similarities than differences, conceptually, between disclosure systems aimed at government information (like procedural or ethics information) and those aimed at the private sector (like nutrition labeling or pollution information).

It seems to me that the disclosures can fail or succeed based on similar issues -- like how the incentives of the disclosers are affected (or aren't affected), and how well the public/press can process and understand the information disclosed.

Perhaps the biggest difference I see between the two is in the reform mechanism.  The Executive Branch is (we hope) remaking their disclosure policies through top-down policies aimed at taking full advantage of new technology, and of a deeper appreciation for the importance of public access.  Congress is similarly grappling with how to create and implement bulk data access.

Private sector disclosure requirements, whether self imposed as a community standard or imposed by the government, seem to me to be much more likely to each be negotiated individually.

If a more coordinated approach to disclosure reform would be helpful, since the same concerns affect many different types of disclosure systems, then it strikes me that pointing out the similarities between government and private sector reform and enforcement efforts may make a stronger case than pointing out the differences.

[kathy and ernie brandon]

I probably shouldn't have bothered but I added a comment to the comment section of this article hopefully providing some kind of counter-weight to what I perceive - and I might be misreading it - to be outright insanity.  The article was great, I thought, and drew some really weird ire from people who made the Grand Canyon leap from the word 'transparency' to 'oh my God she's talking about Communism and now they're marching us into camps.'  I think I was respectful about it, but if I said something wrong I apologize in advance - my comment is signed "Herbert Woodrose".

I think the article states a pretty plain truth, if I'm reading it right:   Government isn't even close to being as transparent as I think we have a basic right to expect.  Not only are they not transparent enough, but their appear to be occasional efforts to keep it that way (putting it mildly).  

So, when a rule or plan is passed, it won't be put into effect or enforced.  And getting rules passed in the first place in this vein are very very hard.

But, at least we have to get them passed, right?  I mean if we don't have it in writing then we have nothing.  At least with writing we can stand there in the cold and whine "but you said... you promised..."  Those comments bother me too from responders:  let's not bother because we don't really think it could ever happen.  So let's just ensure it won't.

But that twisted logic about the nanny state really disturbs me.  On the one hand a commenter takes pride in his intelligence and American Ingenuity and Independence, and decries Ms. Graham's obvious (sarcasm) call for the Nanny State.  But oddly, what she was asking - again, I might be misreading you here - was for Government to in fact STOP viewing us as children with no need or right to know.  To STOP selling us fairy tales and Easter Bunnies and start having the real honest conversation with the adult population that we expect.  

I'm not as smart as some of the commenters on the Boston Globe, which is why I want every shred of paperwork I can possibly analyze, and I'll stay up all night reading if I have to, and frequently do, endless pdf's and essays... but in my layman's analysis this was a call to be treated like adults, and I thought... gee I thought it was pretty tame actually.  I mean no offense by that at all - but it didn't sound like you were calling for anyone to take to the streets or anything.  You didn't use the word douchebag even once.  

I was really surprised at the level of mostly negative responses to the piece.  Either we shouldn't try because government can't be trusted to ever tell the truth - so why try to enforce that they do - or we shouldn't try because the very request that government share the documents and analysis on what is happening to us in this country is akin to communism.  And someone else thought we should distrust everyone also, and only think about a list of memes - literally a list.  

But alright, even if you are only getting the government to release analyses of deregulation, or macro/micro economics, and even if what they release is nonsensical lying - is there really no value in that?  I personally get a great lift when I can catch my government purposely misleading or lying.  Because when you point this out to people, and enough people start saying "what gives?" then the government can't really keep doing that.  It's obviously apathy that allows that to go on.  If we really forced the government to constantly release information and analysis, we could force them to get better at it.  Then we could really get around to -if they won't do it - putting it all in language that is understandable.  You could have a website for example that acts almost like a Google Translator, reducing legalese and boilerplate into basic English, so that elderly black women can't be targeted by mortgage retailers.

Beyond that - and this is really the dream I guess - you could get government actually interested in doing that.  Interpreting documents and plans.  On the very mortgage, underneath the legal nonsense, could be a simple box from the government saying "What this means is you're going to owe____ in the last month of the mortgage."  Or "What this means is your interest rate can change to anything up to and including _____% at ____ date."  I'm not sure why that's insane, or what Nanny State or Stalin have to do with that.

Anyway - these are the thoughts I have from my understanding of the article.

This was a well reasoned and respectful piece, good work.  I think.  Unless you are really advocating barter, communism, fascism, or One World Government including an Amero.  In which case - I really could have used someone explaining all the code to me that my much smarter colleagues at the Boston Globe apparently caught.  So either way I think I agree with the stuff about transparency and government aid.  Just not with your call for a return to Slavery. 

/wink

--
In a time of universal deceit, telling the truth becomes a revolutionary act.
-Orwell

[John Wonderlich]

Here's an interesting somewhat related take, from the Guardian:

http://www.guardian.co.uk/technology/2009/apr/29/cory-doctorow-police-transparency

...

We cyber-liberties types are very big on government transparency – on the right to carry our cameras into every altercation with authority and to put it all online. We make the problems visible, hoping that this will solve them. Little brother watches back!

...

And yet ... nothing much has happened in the intervening eight months. Simply knowing that the police misbehaved does nothing to bring them to account.

Transparency means nothing unless it is accompanied by the rule of law. It means nothing unless it is set in a system of good and responsible government, of oversight of authority that expeditiously and effectively handles citizen complaints. Transparency means nothing without justice.

[Thomas Lord]

On Tue, 2009-04-28 at 20:33 -0700, TPP Weil wrote:

> We make a distinction between transparency policies created to enhance
> the public's right to know about their government from specific
> policies that seek to achieve a particular goal.

[...]

> Are there discussions of some of these
> focused transparency policies
> going on within the administration / agencies?

There appear to be many separate discussions,
all focused on a narrow area. Examples include
advertising regulations in the Credit Cardholder's
Bill of Rights bill and the tossing aside of "mark
to market" rules. At he municipal level, ordinances
that add requirements for restaurant menus are
kind of a fad.

> Are other groups pushing in similar directions to expand the
> transparency agenda to include improving these policies?

I don't know of any that are taking an overarching
view of transparency activities in general. I think
you might have a good, original idea there.

> Looking forward to your thoughts and reactions.

Rather than simply "directing agencies to talk
to one another" (a strategy I suspect leads nowhere)
and end-game goal here might be the legislative
creation of an "office of transparency services"
combined with statutory reporting measurement and
requirements about the costs, compliance, and efficacy
of all transparency rules and regs. The agencies
can talk to that new oversight agency as a hub:
transparency about transparency programs.

I have something else to add but it is less confusing
to put it in a separate message of its own.

-t

[Thomas Lord]

On Tue, 2009-04-28 at 20:33 -0700, TPP Weil wrote:

> Looking forward to your thoughts and reactions.

I think the language you are using (e.g., "focused
transparency") is confusing. Thinking about that,
I came up with a proposed list:

"The Elements of Transparency"

These terms apply to all transparency efforts, public
and private sector -- they are the names of activities
that comprise transparency efforts. It's
a little bit tedious to start off here with
a taxonomy of "elements" like this but I think
you'll see the utility of it afterwords:

"inspection" - measuring data. Regulatory
questions: Who must and who may inspect? Who
must and who may be inspected? What
measurements are taken? What are the limits
of the rights of access of inspectors? (Fill in
answers to those questions for each instance of
a transparency effort. E.g., the SEC has limited
rights to inspect a firm's books; physicians
must inspect for reportable diseases; etc.

"reporting" - conveying the measured data in a
proscribed format. The parameters are who must
and may report to whom?

"aggregation and analysis" - referring to
aggregation and analysis of reports in manners
mandated by the transparency process. Parameters
are who must and may do this and what forms
of aggregation that may or must do or not do.
E.g., the Census bureau must aggregate and analyze
census data on a certain schedule but is strictly
limited in the kinds of analysis it performs.

"disclosure" - the making available of inspection
reports and aggregated data and analysis.
Parameters include to whom data may or must be
disclosed and on what terms. For example,
the "President's Daily Briefing" is a form of
disclosure from the intelligence agencies to the
executive, limited in who may receive a copy.

"visualization" - the informatics (design,
user interfaces, placement, etc.) of
rendering disclosures for the convenience of
various purposes. The parameters here are
what technology of visualization is employed
and why.

"labeling and notification" - is the obligation
to present a proscribed visualization of
certain disclosures in particular circumstances.
The parameters are who must perform the labeling
or notification, what data they may or must convey,
what visualizations they may or must use, and
when or where they must do this. Example: the
nutritional labels on food are required labeling
using a proscribed visualization of a mandated
report about the product.

and finally:

"validation and efficacy testing" - is the
activity of testing the accuracy of inspections,
reports, and disclosures, conformance to requirements,
and the efficacy of a transparency effort at achieving
its stated purposes.

Those are "The Elements of Transparency".

So, what is this good for?

For one thing, some of the distinctions you
are drawing are easier to explain:

For example, the open government movement has
focused largely on transparency processes which
require an agency or office of government to
to perform disclosures to the general public of
data which either is already aggregated internally
to government or which can trivially be aggregated.
And the movement has concentrated on creating
privately held and operated tools for visualization
and notification to try to make this data more
accessible and useful to the general public.

And, indeed, the open government has focused
mainly on those aspects of transparency: disclosure,
visualization, and notification. It has not had a
whole heck of a lot to say about inspection,
reporting, aggregation and analysis, or validation
and efficacy testing.

Yet our "elements of transparency", as the essay
you linked to point out, apply quite as well
to something like food labeling or truth in lending
requirements. Only in those domains, the main current
dysfunctions seem to be inspection, reporting,
and above all validation and efficacy testing.

When we break things down that way I think
your question comes clearer: is it proper to
divide up the elements of transparency and
work on them separately, in niches like that -
or is there also a missing role here that looks
at "transparency in general" - namely trying to
better understand the relationships among the
elements and develop general purpose tools,
techniques, design patterns, and so forth
as a means of improving all transparency efforts
at once?

The question almost answers itself: of course
there should be such an overarching project.
To my eyes, it looks like a very interesting
multi-disciplinary project:

Consider, for example, the set of "who" parameters
in the definitions of the elements. The values
filled in for "who" define a kind of social graph
of information flow and we can use social network
theory math to evaluate the quality of "validation
and efficacy testing" mechanisms. So there's math
and social science to do there.

Or technology: We're into an age now where the
need for "toolkits" (software programs, mostly)
for interactive visualizations are needed to
make our society more agile at processing transparency
data sets. The financial industry has had very
flexible charting tools for a long time and they
get better all the time, for example. Well,
why aren't there free software tools for that
functionality? The same major software components
of *those* systems would be useful for making sense
of, say, public school performance reports or
for tracking categories of "stimulus spending".
So there's some interesting computer science and
software engineering to do there.

Validation and efficacy testing suggest roles
social scientists and political experts.

Inspection being a topic on the rise, there is
a need for legal scholarship ideally able to
discover very general guidelines about the
regulatory options and limitations of inspection
rights.

So I think (at least if you add in my "elements
of transparency") you (or we) have invented a serious
new multi-disciplinary field of study - at least
tentatively - which now waits merely to be populated
by scholars :-)

The "validation and efficacy testing" aspect
of all of this seems to me what Congress should
be very interested in. They pass many, many
different laws in many areas in which these laws
give rise to new transparency processes, each with
a particular set of ostensible purposes.

> Are there discussions of some of these focused transparency policies
> going on within the administration / agencies?
>
> Are other groups pushing in similar directions to expand the
> transparency agenda to include improving these policies?
>

Perhaps it is not enough to have each of those
processes individually make reports back to Congress
but, rather, in the spirit of "transparency about
transparency" the reports should in the first instance
be collected and analyzed by a new Office of Transparency
Services charged with representing the Congressional
perspective on validation and efficacy testing and
on disseminating to reporting entities tools, information,
and general guidance about how to achieve good
conformance and good implementations as inexpensively
as practical.

-t

[Thomas Lord]

(Sorry about the stray "paste" in that
message. The non-sequitor bit towards the end -
the inexplicably repeated quote:

> > Are there discussions of some of these focused transparency policies
> > going on within the administration / agencies?
> >
> > Are other groups pushing in similar directions to expand the
> > transparency agenda to include improving these policies?

was the result of my hair-trigger mouse button
going off and doing a "paste" as I moved the
mouse to hit the send button. Time to get a new
mouse, I think. :-)

-t

[kathy and ernie brandon]

I envisioned something similar, an Office of Transparency or something filling that role.  These ideas are only as good as they are enforced.

Right now most of these functions are handled by separate agencies for separate fields.  Would a single office make sense?  Is it plausible?

I agree that a lot of this is going to necessarily (I think) revolve around software. That has an unfortunately exclusive effect but it is better than where things are right now.  

If there is an Office of Transparency - and this might be too much a of a side-issue - how are those positions staffed?  

Is transparency something we implement in a way that changes the way information is recorded, stored, in the first place?  Do we start the process from the top?  Or is it a matter of creating a body that collects the material after it has already fallen, playing catch-up, or clean-up duty.  Is this a system with penalties?  

Is this Office of Transparency something like the IRS?  Performing random audits to determine compliance, and regularly writing/implementing the codes for practical application of the system?  Or is it more of a benevolent organization, really there more for assistance and facilitation?

We get really quickly into 'who' is going to accomplish this (I personally think it is a job best spread around, so that everyone has a small share in it) and what are going to be the incentives/penalties.  This is going to - if I understand this right - necessitate changing the way the job of Congress is performed in terms of recording and storing data. 

But the big issue I'm seeing is the user experience.  The finance community records things meticulously - but how that data is read, and understanding that there are places within the recording itself where a company is allowed to bend and twist the rules, is a tough job for a governmental office to implement.  How would government explain "goodwill" to the simple investor in a meaningful way that doesn't scare the corporate community?  Things like this are going to be negotiated individually.  But I still think progress has to be made in that direction because it is the only way that these facades can ultimately be broken down.  Once language is demystified, I think the processes that were being covered by that language may have to go away, or find some new, more creative haven.

[Christoph Berendes]

Some of the responses to the article inspire the following

Wholesale transparency: Much of the Transparency movement centered
around Sunlight focusses on "wholesale transparency" - pushing
government agencies to provide timely feeds of data in standardized
formats - and leaving the slicing, dicing, and formatting, i.e. the
decision support, to third parties.

Retail transparency: The article advocates for what one might call
"retail transparency", e.g. regulated formats for comparing different
mortgage loans, or nutritional labeling, which generally bundle both
the data disclosure and the decision support.

Each side could inform the other.

How should "wholesale transparency" be tuned to support "retail
transparency"? For instance, is there feedback from developers using
the rich array of feeds developed or highlighted by Sunlight back to
Sunlight and to government data providers? (I suspect there is, and
that a write-up of lessons learned would be fascinating)

What can those focussed on "retail transparency" learn from current
"wholesale transparency" approaches? For instance, if credit card
companies were required to make the terms for their various credit
cards available in well-documented and "mashable" feeds, Consumers
Union (and the American Bankers Association and the NYTImes, etc.)
could then develop their own "retail" formats and decision support
tools. (Note that I'm not arguing that credit card cos should do this
instead of their current "retail transparency" requirements, rather in
addition. I do think that some competition at the "retail
transparency" level would be a good thing.)

On Tue, Apr 28, 2009 at 11:33 PM, TPP Weil <dav...@bu.edu> wrote:
>