Openness and transparency are good things and the more we have of them the better. Right? Wrong.
In contexts of too little openness – “hypo-transparency” – ICTs can help bring greater transparency, with positive developmental effects. But in contexts of relative openness, ICTs are ushering in a hyper-transparency that will destroy public institutions. As summarised in the figure below, I therefore propose an inverse-U relation between e-transparency and various measures of political development, such as trust in public institutions.
As an experiment, try the following. View your beloved from a very far distance. They are a tiny speck, and you feel nothing for them. Now move closer to view them from a few feet away. Likely you will see much to admire and feel a warm glow (if not, it may be time for an upgrade). Now get up really, really close and examine them in minute detail – take a look up their nose, in their ears, inside their . . . well, you get the idea. That glow’s probably not quite so warm now, is it?
Something similar happens with ICTs and transparency. Applied in corrupt, opaque, self-serving environments, ICTs have been shown to reduce corruption and improve the efficiency and equity of practice. But applied further in democratic environments where a reasonable degree of e-transparency and openness already exists, ICTs can make things worse rather than better.
Through greater e-transparency, ICTs help us know ever-more about the behaviour (decisions and actions) of those within public institutions. The majority of that behaviour will be appropriate. But humans are flawed, so they will always make mistakes, act selfishly, and do bad things. Absent other effects, the greater the transparency, the greater the absolute amount of such inappropriate behaviour that will be revealed, and the less citizens will value and trust public institutions.
Any effects of transparency in reducing the amount of behaviour that is inappropriate are mitigated both internally and externally. Internally, transparency pushes institutions to spend increasing time on non-value-adding defensive activities. These include trying to second-guess and avoid what might cause offence or other negative public reaction; excessive caution in behaviour to avoid risk or failure; and inefficiencies in protecting necessary confidential interactions – the “safe space for genuine deliberation” – from external gaze. Yet, “without the exchange of confidences, it is not possible for people to have real confidence in their colleagues and in the organisations that employ them”[1].
Externally, ever-greater flows of e-transparency data undermine public institutions because of . . .
- Cognitive deficits: the greater the flow of data, the lower the absolute availability of knowledge and motivation among the public to properly interpret that data, leading to a dominance of simplistic interpretations, many of which are negative because of . . .
- Cognitive bias: the negativity bias that causes humans to attend more to bad than good news, to remember bad more than good news, and to form negative stereotypes more quickly which are more resistant to disconfirmation. And the tendency, for example when searching online, to attend to extreme rather than average data. Extreme and negative interpretations of data on public institutions become more prevalent because of . . .
- Political incentives: attention and profile online accrue to those who posit more extreme views, and there are plenty of commentators who have political or economic incentives to criticise current public institutions and who – within already-relatively-open contexts – are able to do so. They have an ability to shape the narrative in part because citizens give up their own interpretation due to cognitive deficits. And thus we have a self-reinforcing spiral.
The impact of this can be seen, for example, in the decline of trust in public institutions in democracies during the Internet era. Dating this from the turn of the century, some illustrations:
- A decline in trust in the European Union from 40% to 30%.
- A decline in Americans’ confidence in Congress from 27% to 7%.
- A halving of Americans’ confidence in news media.
Of course e-transparency is not the only factor behind trust, but a review of some key literature finds little evidence that transparency builds trust. Instead, “in a number of cases, the evidence points in another direction: that is, transparency may ultimately decrease trust”.
This has a number of negative knock-on consequences if lack of trust leads to calls for greater transparency which leads to a further erosion of trust. With only a minority – sometimes a small minority – of citizens trusting institutions, those institutions are weakened in their ability to defend the public realm and public interests. And we see a shift in power from public to private institutions, and from centrist to more extreme political views and parties.
Is this an argument against e-transparency? It is not. But it is an argument that:
- We are guided by the inverse-U curve to give highest priority to using ICTs to open up the most-powerful, least-transparent institutions. That means authoritarian regimes and transnational corporations. Oh, and FIFA. Don’t applaud Edward Snowden until he exposes the workings of the 3PLA, or Julian Assange until he leaks the tax avoidance plans of global IT firms. If you want a transparency hero, pick Herve Falciani.
- We place greater emphasis on accountability than transparency. Transparency, in Furedi’s words, fosters “a political culture of voyeurism”. Accountability – at least when properly designed – fosters reasoned, considered checks and balances against abuse of power.
- We accept there are limits to openness, and that we want transparency but not hyper-transparency: “A democratic society should understand that it is important to uphold the right to the private exchange of views and that not everything officials do ought to be visible to all”[2].
[1] Furedi, F. (2011) Let’s stop kowtowing to the cult of transparency, Spiked, 5 Oct http://www.spiked-online.com/newsite/article/11140
[2] Furedi ibid.
ICTs for Development 
Transparency is a directed and specific form of openness. Problematizing and endless nuancing of the former only helps to obfuscate and distract from meaningful progress towards the latter. Both transparency & openness are only instrumental and not end values. Measuring transparency without adequate care for specific context, by treating it as an end value – in and of by itself – is problematic, although rampant in academic literature. Instead it should be measured indirectly using as a proxy, the effects of reduced opacity. Besides the above issues with construct validity, research thus far also does not explore the causal pathways through which transparency reduces trust.
Ultimately we end up mischaracterizing the concept through imprecision and unwillingness to come up with more plausible causal explanations. The plea to emphasize accountability is even more dubious, given that it is an even more magical concept than transparency. Finally instead of pushing back against openness, scholars can lay out a limited legal justification for government secrecy, which can allow the principled use of political incentives. Similarly ethical guidelines for dissemination of public information can overcome the issues with cognitive deficits. The way to reduce the effects of cognitive bias in democracies is for political actors to actually work towards reducing extreme bad news rather than trying to put the genie back in the bottle.
Some guy, absolutely convinced that your words treats what I want to speak. We need context to talk about transparency and it is not the end but the mean to achieve something, for example the accountability or the openness (open government?).
Hi Richard, Great to see you writing about this. It is something that has been on my mind for some time. Similar to “some guy” I think that openness requires a more nuanced analysis recognising that transparency is only one element of it. I try to unpack openness along the same lines that Jonathan Haidt unpacks morality in The Morality of Openness. That analysis led me to shift my focus from openness, which is a means not an end, to the issue of trust. In The Future of Open and How To Stop It, I come to the conclusion that the monotone lens of openness has actually shifted the focus away from where it ought to be. And now I can’t help but see everything through the lens of trust.
This is a very interesting discussion. One that has surfaced time and again as I have worked in the field of e-democracy, open gov, big data and the implications these have for developing regions. A key question that emerges from this is: Do we see a recurring pattern here? Buzzwords come and go (transparency, open, participative, co-creation, inclusion etc). But are there certain inherent mechanisms in place to protect, entrench and sustain the status quo of power hierarchies? Are we really challenging the political mythology of tech assessment, when we look beneath the surface of such deceptive terminology?
I have come to see the whole ecosystem in an evolutionary and adaptive sense. As new technologies become available, new strategies emerge to wield them to certain intrenched interests/advantages. This is not to say access and capabilities are hugely transformed by these innovative technologies and ideas, but the underpinning assumptions of equity, freedom, justice, a better life for all….remain contested.
From a discussion of this on LinkedIn, one useful emergent point: how would we define “hyper-transparency”? Despite some modest currency as a term, it’s hard to find a definition so let’s try to create one. We can derive some components from what’s been written to date:
– it is associated with ICTs; specifically with the advent of social media/Web 2.0 (e.g. Goodman 2011: Corporate communications in times of stress; Bonime-Blanc 2014: The reputation risk handbook);
– it is an existent state (at least, all sources discussing it make this assumption), so it falls below hypotheticals such as a condition of absolute openness of all things; and
– being ‘hyper’ it is above and beyond the norm. This is the really difficult part: how do we define normal levels of transparency when these vary from context to context, and from time to time? Yet at the same time a definition can’t be totally relativist – if, say, North Korea experiences an increase in transparency from the norm, it does not become hyper-transparent.
So, first draft: “A level of open availability of data on human behaviour facilitated by digital technology and significantly beyond the norms of late 20th-century Western democracies”.
Very very interesting. Thanks a million. But can’t we extend this definition to non-democratic societies if only include the level of people’s awareness and capability?
For instance, one can exemplify Iranian people, especially after the 2009 defeat and argue that how Web 2.0 created “a political culture of voyeurism” or negativity amongst people.
The idea is certainly supposed to be global – I would have to leave it to Iran experts to place the country on the curve, though I’d be surprised if the argument was that Iran had become too transparent. But perhaps the argument is that the curve works differently in different parts of the political sector; some parts can be insufficiently transparent while others are too transparent?
Very useful set of papers relating to this topic, suggested by Savita Bailur:
– Bannister, F. and Connolly, R. (2011). The Trouble with Transparency: A Critical Review of Openness in e-Government. Policy & Internet, 3(1), pp.1–30.
– Bauhr, M. and Grimes, M. (2014). Indignation or Resignation: The Implications of Transparency for Societal Accountability. Governance, 27(2), pp.291–320.
– Costa, S. (2012). Do Freedom of Information Laws Decrease Corruption? Journal of Law, Economics, and Organization, 29(6), pp.1317–1343.
– Grimmelikhuijsen, S. (2012). A good man but a bad wizard. About the limits and future of transparency of democratic governments. Information Polity, 17(3), pp.293–302.
It’s been pointed out that one ought to love one’s beloved even at a very far distance, so I should clarify that I meant in my analogy so far away you couldn’t recognise them. Alternatives suggested for an inverse-U were viewing a painting or a football match.
McKinsey on the dark side of (organisational) transparency: http://www.mckinsey.com/business-functions/organization/our-insights/the-dark-side-of-transparency