Is open data at odds with citizens’ privacy?
As governments collect more data about their citizens,
new privacy dilemmas arise
by Shaida Badiee,
Managing Director, Open Data Watch
Co-Chair, SDSN TReNDS
Open data isn’t a new concept. It’s been a conversation-starter in government circles for years, as we’ve seen more and more governments launch open data initiatives in countries, states and cities across the globe.
But beyond the PR opportunity and desire to appear open, has data openness actually made a difference?
Thankfully, the answer is a resounding yes. The evidence shows how both citizens and governments have benefited, as governments continue to open up their official statistics and datasets for public review.
The value of open data
For example, in 2016 when the government of Paraguay looked for new ways to prevent the spread of dengue fever, the national health surveillance department opened up their data on dengue morbidity. Researchers were able to use this data to create an early warning system that can detect dengue fever outbreaks in every city throughout the country at least a week in advance.
In Namibia, the government turned to the country’s largest telecommunications provider to identify where citizens were at high risk of contracting malaria, using open mobile data combined with satellite data, enabling the Ministry of Health to target and distribute 1.2 million bed nets to the most vulnerable communities.
Palestine too has made great strides in opening their data, through only modest website and policy changes |
And in Uganda, since 2015, open data has been instrumental in efforts to prevent banana bacterial wilt (BBW), a crippling banana disease that had destroyed nearly 90% of crop yields on many farms and resulted in a national economic loss of approximately $360 million per year. The Ugandan government used open data from U-report, a community polling project, with vital information on BBW’s disease patterns, that the government used to disseminate information about treatment options and actionable crop-protection strategies to some 190,000 citizens.
And it’s not just in the public finance, public health, and agriculture sectors that open data can make a difference. It can also foster greater citizen engagement and enable collective problem-solving, resulting in the more equitable and democratic distribution of information, an increase in public trust, and a greater ability to expose corruption.
Take the case of Nigeria, where the civic start-up BudgIT has exposed 41 million naira ($113,000 USD) in fake and frivolous spending in the federal government, which resulted in a public outcry about pay levels and a public servant salary cap. In Burkina Faso and Indonesia, the availability of open and accessible electoral data in recent elections has continued to strengthen trust between citizens and elected officials.
A long way to go
According to the 2018-19 Open Data Inventory (ODIN), which assesses the coverage and openness of official statistics in nearly 180 countries, 11 countries have seen data openness increase by more than 35% in the past two years. Yet the median ODIN country score is still a mere 41 out of 100.
For some countries, the reason for this is a lack of political will and insufficient investment in their national statistical office (NSO) to enable them to modernise, regularly update, and publish their data. For others, it’s technically challenging — they lack the e-infrastructure to upload their data and the software to make it consistently accessible and machine-readable in a timely manner.
There aren’t any easy solutions for the protection of privacy in microdata, and there is a need for more regulation and well-designed processes to prevent the release of citizens’ personal data. |
Yet, these modernisations don’t have to cost vast sums or be overly complicated. For example, in 2017-2018, Sierra Leone’s NSO launched a new website and adopted more open, although still limited, terms of use. As a result, coverage of subnational data increased, and the new website has made official data and metadata easier to find. Statistics Sierra Leone is a great example of how simple but effective website updates can lead to improved accessibility without the expensive infrastructure required to build complex data portals.
Palestine too has made great strides in opening their data, through only modest website and policy changes. Since 2015, the Palestine Central Bureau of Statistics (PCBS) has made a conscious effort to release more of their data in machine-readable and non-proprietary formats, to make available more metadata, and has also created new terms of use that are fully open.
Data and privacy
Much of the open data that governments share come in the form of aggregate statistics or “macrodata” — for example, the national share of individuals below the poverty line. This macrodata relies on microdata in the form of individual or household-level data, which NSOs collect through censuses and surveys. While microdata is potentially very valuable and essential for the creation of many indicators, they can and have been misused.
There is still an ongoing debate about how and when microdata collected and managed by governments can be released, and how governments and the private sector can and should use it. Furthermore, there is still no consensus on how much protection is needed to maintain the confidentiality of microdata from surveys.
Governments have a responsibility to use this data ethically and legally and to prevent disclosure of personal information. Therefore, they must balance citizens’ “entitlement to public information” with their right to privacy. This balancing act is exemplified by the recent conversation surrounding “differential privacy” in the ongoing US Census 2020. Using statistical techniques, officials are adding errors to individual census results in order to prevent the reidentification of individuals while preserving the statistical snapshot of America.
The trade-off is that while individuals’ identity might be protected, an analysis using the microdata may now show incorrect tallies for very small regions or groups.
Striking the balance
There aren’t any easy solutions for the protection of privacy in microdata, and there is a need for more regulation and well-designed processes to prevent the release of citizens’ personal data. Governments must maintain secure control of the data they possess, and accidental or intentional releases of personal information should be taken very seriously.
Fortunately, several governments have begun to put into place these necessary policies and legislation. Just last month, for example, Kenya’s high court suspended the country’s new national biometric identity program until the government enacts laws to improve its data privacy and prevent discrimination against minorities.
This continues to be a balancing act, for while laws on data privacy are a step toward building a framework for the control and responsible use of microdata, they can also go too far, stifling innovation by preventing legitimate uses of data. But in other cases, such as biometric surveillance, they may not go far enough depending on their mandate and implementation.
The challenge ahead is how we balance the availability and use of microdata with the privacy of citizens and build trust in governments to guard and use information safely and securely. |
As countries adapt to the digital age, data openness and protection will be central. The challenge ahead is how we balance the availability and use of microdata with the privacy of citizens and build trust in governments to guard and use information safely and securely. We should continue to build on the evidence we have that demonstrates the benefits of open access to official statistics for both governments and citizens.
Governments can strike this balance by adopting data governance frameworks that provide clear rules about how macro- and microdata should be treated throughout the data value chain. With this perspective in mind, good data governance can promote innovation, efficiency, sustainability, growth in use, and increase the value of data, while building trust and safeguarding against the misuse of data.
(This article originally appeared in Apolitical)
