OpenGov Conversations: Greg Michener on Creating Effective Transparency Policies


This post is part of our series, OpenGov Conversations, an ongoing discourse featuring contributions from transparency and accountability researchers and practitioners around the world.

robert_gregory_michenerThis post, by guest author Greg Michener, responds to the following questions:
What are the three most important conditions that make transparency policies effective? And must these conditions precede effective transparency interventions, or can transparency policies actually positively bring about these conditions? Read the other responses here.

Transparency is a slippery concept but important enough that it should be handled with some degree of precision. Unfortunately, the concept is often stretched out of shape and credit-hungry policymakers adopt transparency policies with little regard for preconditions. Here I’m going to take a few positions on preconditions, and my underlying point is that transparency is highly contingent. In order for transparency policies to work we need to take a more cautious approach to conceptualizing ‘transparency’ and to understanding the incentives of supplying and demanding transparency.

Is it really transparency?

I’m going to start out with something basic, a framework that can help unmask bad information masquerading as ‘transparency’. For transparency policies to be effective, they first need actually be transparency – and optimally, high quality transparency. Necessary and jointly sufficient conditions for the existence of transparency are visibility (completeness, findability) and inferability (the ability to draw accurate conclusions).

You might hear people talk about how a process was ‘transparent’. If no record of that process exists, however, it does not fulfill the first dimension of transparency – visibility. Visibility is all about information: completeness and findability (can you find a needle in a haystack?).  Inferability works a little differently; it depends to a large extent on the intended audience and it grows more effective as more attributes are added: verification by a third party, a simplifying heuristic (e.g. visualization), disaggregation, and machine readability are the primary ways of increasing the inferability of information and the quality of transparency.  These ideas about transparency as a concept come from a paper I wrote with Kate Bersch and that will be coming out in the August version of Information Polity. Kate and I see visibility as mostly demand driven and inferability as mainly supply driven.

Are there good incentives to supply and demand transparency?

It’s often said that ‘leadership’ is required to ensure effective transparency policy. There are many leaders, however, whose best intentions are crushed by transparency-adverse allies. Ironically, it is often the leaders with the greatest power to advance transparency – such as those who control the legislature – who are least likely to do so. The point being, while leadership is necessary, it is contingent on institutional incentives. The demand side is just as important. If no one is looking for good information – i.e. there is no muckraking, public interest news media – then leaders will have few reasons to supply effective transparency, and transparency will serve little purpose.  So agency (e.g. media or citizens) and institutions (e.g. the legislative situation) have a lot to do with the effectiveness of transparency policies.

The Devil is in the Details

When we talk about transparency policies, we are speaking about two types of policies, those where transparency provisions serve as lynchpins, such as political finance regulation, and transparency policies proper, such as freedom of information laws (FOI).  The effectiveness of both types of policies is contingent upon small details. Drawing on the first example, countries like the Dominican Republic and Panama have political finance regulation, but they do not make campaign contributions public. Instead, Electoral Tribunals vet information provided by political parties and then provide summary verifications to the public. Lacking the visibility and inferability of source information, information suppliers raise too many questions: is information being gamed? Are electoral authorities doing their job, or are they captured by partisan interests? When the supply of transparency comes into question, demand withers. The Institute for Democracy and Electoral Assistance (IDEA) provides a list of over 40 political finance regulations. If number 38 – are political finances made public? – is answered negatively, the utility of the entire political finance regulation apparatus begins to fall apart. In this sense, transparency clauses serve as ‘lynchpins’ that hold together policies and policy structures.

My second example is freedom of information (FOI) laws. Here, the ‘lynchpin’ conditions are small details that can virtually neutralize the effectiveness of a law. For example, if applicants must justify their requests to authorities, as in Ecuador or Jordan, the entire effectiveness of laws must be called into question. Seekers of public information should have to justify themselves no more than those who practice any other fundamental right, such as those who freely express themselves, assemble, or petition.

Does your Advocacy include Prescriptions and Proscriptions?

The example of FOI brings up a point that needs far greater attention. When advocating for transparency policies – and the details that make or break them – advocates can take two approaches: ‘prescriptive’ or ‘proscriptive’. Prescriptive advocacy suggests to policy adopters (e.g. politicians) what they should do; proscriptive advice is what they should absolutely not do. It hardly needs to be said that prescriptive advice far outstrips proscriptive advice in the marketplace for transparency policies. In my view, this is a big mistake. As the above example on justifying FOI requests should convey, advocates should identify a list of absolute no-nos. ‘Thou shalt nots’ are rarely explicit enough.

The Center for Law and Democracy and Access Info’s Right to Information (RTI) Rating, for example, subtracts only 2 out of 150 points if citizens need justify a request. What type of message does this send? The RTI Rating provides an excellent prescription of what a law should look like, but it does do enough to stress what a law should not look like.

The RTI index is not alone. One of the reasons so many organizations fall down on ‘rating and ranking’ transparency is that they aim for ‘unified’ indices as opposed to disparate indicators. The problem is not just that they under or over weight certain provisions, it’s that certain provisions are so critical they should be part of a separate ‘proscriptive’ set of indicators; a list of ‘thou shall not’ commandments that can be observed side-by-side with the traditional ‘thou shall’ prescriptive index. Infringing any of these commandments should be viewed as putting at risk the effectiveness of transparency mechanisms.

The need for ‘proscriptions’ is elementary to encouraging desirable behavior. We might tell a child “wash your hands before dinner” (prescriptive), but some advice must be proscriptive: “never get into a car with a stranger”.  Whereas other policy domains may have few ‘do nots’, there are certain absolute no-nos when designing transparency policies. We need to make these clearer.

So all in all, my take on transparency policy is that preconditions are key. We need to pay greater attention to transparency as a concept, and to promoting adequate incentives for good quality transparency (supply side) and its demand (e.g. free and independent media). These incentives must also be incorporated into our own advocacy, which must be both positive (prescriptive) and negative (proscriptive).

Bio: Greg Michener is a Canadian citizen, naturalized Brazilian and professor at the Brazilian School of Public and Business Administration at the Fundação Getúlio Vargas.. Michener is currently working on a book to be published by Cambridge University press, on determinants of robust transparency and freedom of information (FOI) laws across Latin America. His research addresses issues such as political transparency, access to public information, the open-data movement, legislative policies, the media as a political institution, the role of the business sector in the policies of good governance and qualitative methodology.