The Internet, Policy & Politics Conferences

Oxford Internet Institute, University of Oxford

Tracey P. Lauriault, Peter Mooney: Crowdsourcing: A Geographic Approach to Identifying Policy Opportunities & Challenges Toward Deeper Levels of Public Engagement

Tracey P. Lauriault, Programmable City Project, National Institute for Spatial and Regional Analysis, National University of Ireland at Maynooth (NUI

Peter Mooney, Environmental Protection Agency of Ireland and Dept. of Computer Science, NUIM



Crowdsourcing in geography comes in a number of models, here we examine three: Volunteered Geographic Information (VGI), Citizen Science (CS) and participatory mapping (PM). These are long standing knowledge generation models to study the human and physical world (Goodchild 2007, Audubon Society 1900, Peluso 1995). The field notes of new world explorers, for example, show that amateur scientists and officials extracted knowledge from those they colonized and used those data to inform colonial policies. Fortunately, a more critical form of geography combined with internet technologies such as Google Maps, APIs, and webmapping among other things such as interoperability standards and geotagging, have democratized crowdsourcing and made it a little easier, transforming members of the public, albeit often a highly qualified public with specialized, scientific and technological skills; from passive consumers of authoritative data into data producers (Dodge and Kitchin 2013) or prosumers (Ritzer 2008).

Successful VGI and CS examples such as OpenStreetMap, FixMyStreet, Zooniverse and Project Noah, are grassroots initiatives. PM, which here implies community mapping, counter mapping, and participatory GIS; includes stakeholders collaborating with cartographers and geomaticians in action oriented research settings to produce maps (e.g. All Island Research Observatory, Cybercartographic Atlases). These three established crowdsourcing models, provide a deep level of public engagement and reliable data, yet continue to inform public policy from the outside. To date there are very few CS, VGI or PM endeavours driven or adopted by government., GeoConnections and the Canadian Geomatics Community Round Table, Bird Atlas of Britain and Ireland, and the Notification and Editing Service (NES) are examples where the potential of crowdsourcing as discussed here are being examined. Researchers and policy-makers have however just begun to examine the opportunities and challenges presented by these data driven models of public engagement at all scales. This paper therefore aims to describe and critically examine these models by applying a data assemblage framework (Kitchin 2012) to identify actors, forms of knowledge, technologies, data quality, practices and norms within a select number of case studies from Ireland, Canada and internationally toward the early development of a typology of practical policies to help government adopt crowdsourcing as a public engagement strategy.

A number of affordances for government to use data collected through CS (Bonney et al 2009), VGI in a national spatial data infrastructure context (Budhathoki et al 2008) and PM (Bryan 2014, Brown and Kyttä 2014) have been identified. This includes the potential for the public to act as “sensors of their environment” (Goodchild 2007, Johnson and Siber 2013), to analyze those data with government (de Leeuw et al 2011) and in the case of local and traditional knowledge, it is an opportunity for other worldviews to be represented such as indigenous toponyms, land use and occupancy, or biodiversity (Stafford et. al 2010, Taylor and Lauriault 2014). As governments in many jurisdictions must act with reduced budgets and limited human resources CS, VGI and PM offer an opportunity to include the public in data collection and analysis toward a more grounded approach to evidence informed decision-making. We suggest that governments can build public engagement opportunities by capitalizing on these three established geographic crowdsourcing models, furthermore, these models may also inform crowdsourcing processes more broadly. 

Governments to date, even with the adoption of open government principles and the launching of open data portals, have been reluctant to embrace crowdsourcing. Performance and accountability measures are of primary concern and focus of public managers and administrators. Governments cannot simply just begin to use crowdsourced data without understanding policy implications. The adoption of crowdsourced data introduces many challenges including the sustainability of VGI, CS and PM; authorship; public officials as public mediators; data and software ownership; and the reliability of data sources, to name a few. The widely held perception that VGI and CS data and to a lesser extent those of PM contain many data quality problems compared to authoritative sources is a major constraint. There are also legal implications for governments if the data they use are inaccurate, biased, or somehow flawed (Scassa 2014). Scientists and data analysts inside government departments cannot be expected to simply begin using crowdsourced data without first developing suitable structures, technologies, policies to do so. Formalized processes for the collection of VGI, CS and PM data (incorporating data quality controls and verification) coupled with the development of government structures to use these are therefore required. 

Concurrently, governments are: signing onto the Open Government Partnership; adopting open government practices and launching open data projects which include increased public engagement. Governments have also recently begun to understand how to leverage myriad internet technologies (e.g. Web 2.0, APIs, GeoWeb) and are participating in new forms of public engagement activities such as hackathons, data dives, app contests and un-conferences (e.g. GovCamps) where officials and the public collaborate to derive meaning from authoritative datasets or mash these up with CS/VGI data into new apps. Interestingly, the data assemblage of open data and open government often differs from CS, VGI and PM. It is suggested here, that there is a need to study the lessons learned arising from successful efforts and demonstrate ways for governments to engage with VGI, CS and PM communities and that the spirit of the open data movement in government provides an opportunity for this form of experimentation. 

This paper will identify some of the common barriers making officials reluctant to adopt crowdsourcing. We will compare these barriers with the best practices identified in three geography based crowdsourcing models and some open data examples, with the objective of developing public policy recommendations to assuage government to include crowdsourcing into their public engagement practices. A recent report by the European Commission (Science for Environmental Policy 2013) remarked that in Europe many initiatives to incorporate CS in environmental policymaking as part of a broader development of participatory forms of democracy are in their early stages. This paper will contribute in a small way to the advancement of knowledge in this area.

Finally, we argue that governmental reform, increased demand for transparency and accountability combined with open government practices are changing the relationship between the public and government and can serve as springboards for developing the concept of data informed co-governance and open opportunities for the incorporation of VGI, CS and PM into government decision-making, policy development, and meeting regulatory reporting obligations.

Tracey P. Lauriault, Peter Mooney