CIRA: For starters, can you tell us a bit about Open North and the Represent API?
Jean-Noé: Open North is a non-profit organization that works to help citizens use technology to keep their elected officials accountable. We built the Represent Civic Information API to help voters easily and quickly find information on their elected officials, such as their names and email addresses. Every election, groups build tools to inform electors about their candidates, and our API is indispensable in determining the elector’s electoral district and in providing information about each candidate.
CIRA: How do you get access to the data that appears in Represent?
Jean-Noé: Interestingly, most of the data made available via the Represent Civic Information API is, in most cases, not open data.
All the electoral boundaries and elected officials’ contact information are collected from government sources, who hold the copyright and choose the license. Where possible, we redistribute the electoral boundaries in the same formats and under the same licenses as the government source. However, there are many electoral boundaries in the API that we do not have permission to distribute; in these cases, our API is the only public source for this data.
With respect to elected officials’ contact information, thanks to funding from the .CA Community Investment Program, we were able to convince over 25% (17) of the municipalities with open data catalogs in Canada to publish their elected officials’ contact information in a CSV format as open data. For the other 77 jurisdictions in the API, our scripts automatically extract the contact information from government websites once a day; the copyright status of this information is uncertain.
CIRA: What are the current barriers to governments who have yet to adopt open data standards in Canada?
Jean-Noé: Thanks to the .CA Community Investment Program, we interviewed three municipalities in Ontario. For two interviewees, there were no major barriers: one now has an open data policy, and another has a working group to move open data forward. Open data adoption seems to be happening later in these municipalities due to their initiatives having started later and having followed a more formal process requiring council approval. In contrast, many early adopters of open data in Canada didn’t follow a formal process and were able to move faster. Many challenges from these interviewees had to do with the need to request permission and receive approval at multiple steps in the implementation of an open data initiative; cutting red tape where possible would help overcome these challenges.
For the last municipality where progress on open data has stalled, the challenge was not that there were strong objections to open data; rather, the challenge was overcoming competing priorities. Better evidence of the impact of open data is perhaps the best way of overcoming this challenge. Unless mayors and councilors are already convinced of the benefits of open data – based on the, frankly, limited evidence to date – stronger evidence is required to convince councils to make open data a priority.
CIRA: What has been the response from the governmental community to this initiative?
Jean-Noé: The response has been very positive. Many government open data initiatives are seeking opportunities to standardize datasets across jurisdictions, and elected officials’ contact information is an easy dataset to produce and maintain, since this data changes infrequently.
The main barrier to adoption is the lack of a central authority within most government open data initiatives. Within the majority of corporate-wide initiatives, each department decides what data to make available and how to make it available; authority is decentralized. In the case of government, the open data lead lacks the authority to require the department responsible for elected officials’ contact information – often the clerk’s office – to publish the dataset or to adopt a specific format for the dataset. Our strategy has been to support the open data lead with a variety of arguments in favor of a CSV schema.
CIRA: How did Represent API begin, and what were the challenges in making something for everyone?
Jean-Noé: The Represent API was started through a few community hackathons in Montreal, organized by James McKinney and Michael Mulley. The project is deeply indebted to Michael Mulley for voluntarily authoring the first version of the software. His original designs have survived relatively unchanged over the three and a half years of operation. The main improvement to meet users’ needs was to offer CSV downloads for users who are unfamiliar with APIs.
For more information on Open North and the Represent Civic Information API, visit http://www.opennorth.ca.