What are the reasons not to open up public data?

Its understandable that there are fears and resistance to open data within government, and whether you are an open data advocate or skeptical about it, its worth engaging with the following common concerns and fears around open data in cities.

Reasons not to open up public data

Response to questions

No one cares about our data.

You will not know that until you have identified different type of data users and developed effective feedback mechanisms to gauge their interest and the potential ways that data could be used and reused.

Our data is not perfect. It’s not even good.

Good is very subjective, especially if that assumption isn’t being expressed by current data users themselves. Data that isn’t perfect still offers very valuable information as long as data quality limitations are stated clearly. Imperfect data is an engagement opportunity.

Couldn’t the data we release be used by people with bad intentions

If data can be requested by a public access to information (PAIA) request, and doesn’t reveal the personal identity, then it should be accessible to the public. For example, public safety must not be jeopardised and in most cases sensitive data can be anonymised or masked to address these concerns.

Someone would need a professional degree just to understand the data.

There are plenty of people with or without professional degrees who would like to use and reuse public data. Having access to public data is a right, not a privilege. Its a great idea to provide access to help users become more literate, and also to ensure that metadata and other documentation is robust and easy to use, however this is not a reason to not open data. More and more people are becoming data literate. There are courses are offered online, as part of regular community meet-ups, and are now integral to university curricula.

This is going to be so much work – I don’t have the time!

We’re all very busy, but there are opportunities to release datasets that are already in good quality that requires minimum processing. This toolkit explains that you don’t have to move from closed to open overnight. Start small, test the waters, engage, build your pilot, explore the demand, and make the case for a larger initiative. One of the benefits of open data is actual reduction of staff time for data requests.

We already sell it…oh, we don’t? Well….we should!

The returns on investment and benefits of opening up data are now well documented (see the benefits of open data). Selling data creates a barrier that constrains innovation and prosperity in your community. It’s important to lower the accessibility bar to benefit various users, like entrepreneurs and social innovators developing their business model, building a prototype, etc. It has also been found that the administrative costs for delivering the data outweigh the revenue.

We don’t know what “they” will find

We simply can’t control the interpretation of datasets. However, like journalists who rely on their credibility, it’s highly unlikely that open data will generate false analysis without being discredited. And, well, what do you have to hide? Transparency is at the core of our democratic system. Many municipalities around the world now routinely provide expenses of their politicians online.

If we give it to them they will just want more.

Possibly! And that’s a good thing. But expectations should be managed proactively. Communicate your plan and let data users know the capacity that you have to meet their demand to avoid frustration or requests that simply can’t be met because of the workload involved at the early stage of your open data initiative. Promotion of Access to Information (PAIA) requests have standard response times, so should open data.

No, it’s mine and you can’t have it!

Really? Open data has changed the relationship between data producers and data users. Data producers or managers are responsible for the data. They are not the owners of public data, which is paid for by taxpayers after all. Government data is a public asset.

That’s what access to information requests are for

Opening up data proactively according to open data principles will decrease the number of access to information request. It’s called proactive disclosure. The key is to make sure that people know where to find the data once it’s released, which is one of the core principles of open data (i.e. discoverability)

What about POPI?

Well, simply de-identify data before sharing, except where required by law, ie, tender awards, appointments, black listed companies. It's great that municipalities are concerned about the privacy of their customers and this concern is adequately catered for in the Protection of Personal Information Act No 4 of 2013 (POPI). POPI refers to the process of de-identification which allows for the release of data provided that a person's or business identity is not connected with any data disseminated publicly. This means that columns containing personal information such as the name of person, identity number, email address and telephone number must be deleted from data before being publicly. POPI also stipulates that data should not be released if it can be manipulated or linked to other data and thereby identifying a person. POPI provides for the release of personal data only when required by other Acts of Parliament, for example, tender awards and valuation rolls. So, open data is not in any way a contravention of POPI, in fact responsible data practise within open data requires we take this into account when opening access to public data.

We don't have open data legislation or policy!

Well, that is not really a problem in terms of getting started with open data. A quick scan of municipalities show a number of them sharing data information as guided by the constitution and national policy and legislation, their own policy environment, and their strategies, goals and commitments. For example, here is some data shared by Ekurhuleni on their open data catalogue, here are GIS shapefiles shared by eThekwini Municipality, and here is where the city of Tshwane shares their tender awards. Open data initiatives really just provide momentum and identity for what is already happening in some municipalities around the country, and provides support and access to resources from the open community.

Don't we need to do a bunch of things before we open data?

Well, why? What things? Given that we already share (or should be sharing) open data on municipal budgets, valuation rolls, city maps, and many others, municipalities are already in place to share data. If you need assistance in upskilling or resourcing to open and share data, get in touch with the Cities Support Programme who can connect you to resources and partners who can assist with this.

Who is going to sign-off on this?

Who ever gave authority to provide open data should sign a formal letter of mandate. The senior manager of the department tasked to serve the open data should sign off that the open data confirms to the standards of open data.

Who is going to manage this?

Ideally the person in the unit or department responsible for data. However, see the entry above on concern about the amount of work open data takes - the aim is to start small and sustainably.

Maybe we could just release non-threatening data to comply

The challenge with holding back data that should be public but that is sensitive or "threatening" is that it inevitably gets out, and at this point the municipality does not only have to deal with this, it has to do so on the back foot, reactively rather than proactively, and has also lost the high ground of credibility and transparency. Proactively releasing data and driving the dialogue on transparency builds trust and creates positive sentiment in citizenry when challenges come along.

This will be a nightmare for councillors if our constituents had this much detail

Once again, one of the current challenges (or "nightmares") for administrators and councillors alike is the amount of misinformation and distorted information that drives public opinion and sentiment. Constituents have this much detail already, it just currently comes from second hand sources, social media, and hearsay. These create the confusion that allow others with alternate agendas to capture interest groups. It is unlikely, and undesirable, to have a perfectly harmonious relationship between councillors and constituents, but open data at allows focus on the correct information and for the issues that need public attention to have it in the frame of truth rather than distortion. This can then lead to true evidence-based engagement and, hopefully, resolution.

*This table and content has been reproduced from the DIY Open Data Toolkit by the Government of Canada (created in partnership with Open North), and, where relevant, contextualised and added to with local South African context. Source: https://open.canada.ca/en/toolkit/diy/2-say-hello-open

Is there a risk you believe to be missing from this table? Please submit it here and we will attend to it and add it to the list!

Still not convinced? We explore more on the risks and challenges of opening data next.