How open data is prevented

The argument from diffuse responsibility is one of the most effective blocking patterns in the open data debate, because there is barely anything to argue against: if nobody is responsible, nobody can be held to account. The passing back and forth between IT departments, specialist units, municipal data centres and federal authorities is often not the result of genuine uncertainty but of decisions about data governance that were never taken.

Questions of responsibility are solvable. Open data officers, data coordination units and governance frameworks, such as the Berlin open data act or the open government data principle in the German E-Government Act, show how institutional responsibility can be created deliberately. Where there is political will, responsibility is established. Where it is missing, responsibility is described as missing.

The argument that publication must wait for perfect, uniform standards has a structural logic that makes it theoretically extendable for ever: because standards can always be improved further, the argument can in principle never be refuted. In practice it is rarely a technical description and more often a political position.

Robust interoperability standards do exist: DCAT-AP.de for Germany, DCAT 2.0 as a W3C standard for Europe, INSPIRE for geodata, OpenAPI for interfaces. Municipalities and federal authorities publish successfully with these standards every day. Pointing to missing or inadequate standards usually masks an unwillingness to begin at all.

Data protection arguments against open data are rhetorically effective because they carry moral authority: who argues against protecting personal data? At the same time these arguments are routinely applied to datasets that contain no personal data at all: budget figures, infrastructure information, aggregated statistics, land use plans.

Article 86 of the GDPR contains explicit opening clauses for official documents. The German Act on the Re-Use of Public Sector Information and the EU Open Data Directive 2019/1024 create positive obligations to open up for the public sector. Data protection is a legitimate argument for data that really is personal; for the large majority of administrative datasets it simply does not apply.

The publish-and-pray pattern reflects a technicist understanding of open data: you make the data available, and then, use happens. That understanding ignores the fact that open data is a sociotechnical ecosystem requiring relationships, trust and feedback loops. Portals without documentation, community building and feedback mechanisms have measurably lower usage rates.

Successful open data initiatives, from GovData to the Berlin data portal to the open data portal of the city of Cologne, invest deliberately in communication, named contacts, APIs and user workshops. They understand publication not as a one-way process but as the beginning of a relationship. Publication without care is not open data; it is digital filing.

The argument that open data harms business interests has some empirical plausibility for companies that have so far profited from exclusive access to data collected by the state. The view across the economy as a whole, however, is clear and robustly evidenced: open data generates value rather than destroying it.

The McKinsey Global Institute (2013) estimated the global economic value of open data at three to five trillion dollars a year. The European Commission puts the EU-wide value at 184 billion euros a year and rising. The example of the UK Ordnance Survey, which earned more from value-added services after opening its geodata, shows that open data creates new markets and business models. What it destroys are outdated rent positions, not innovation.

Scenarios of misuse are rhetorically effective because they address real risks and because they are hard to falsify: a plausible case of misuse can always be constructed. At the same time the analogous risks of closed data are routinely ignored: data leaks, corruption inside authorities, uncontrolled commercial trade in administrative data.

The Open Knowledge Foundation formulated precisely in the Open Definition that openness does not mean defencelessness. Licences govern the conditions of use, those conditions are enforceable, and the wider public as a body of scrutiny is often more effective than institutional gatekeepers. When everyone has access, errors, manipulation and misuse can also be identified and corrected faster. Transparency is frequently the better protection.

Cost arguments appeal to real budget constraints and are therefore politically hard to counter. At the same time the actual cost of open data is regularly overestimated and the savings structurally ignored: fewer freedom-of-information requests, fewer duplicate surveys, better data quality for internal use, reduced maintenance effort thanks to external validation.

Houghton (2011) showed for Australia that the social returns on open research data exceed the costs many times over. EU programmes such as the Connecting Europe Facility, Horizon Europe and the structural funds co-finance open data infrastructure. And in many case studies the actual cost per published dataset lies far below the original estimates, particularly where data curation is built into the collection process from the start.

Fears of losing control are understandable. An institution that has controlled its data for decades fears the consequences when others work with it and possibly reach different, more critical conclusions than the institution holding it. This fear is rarely stated explicitly but is often structurally effective.

The open data ecosystem offers differentiated instruments of control. Creative Commons licences (CC BY, CC BY-SA, CC BY-NC) govern the conditions of re-use precisely and are legally enforceable. Embargo periods protect one's own research and competitive interests. Tiered access, fully open, registered, on application, allows openness in stages. Persistent identifiers secure authorship permanently. FAIR rather than merely OPEN is not a step backwards but a differentiated practice.