Assessing the Public Value of Analytics: Supporting Infrastructure Investment Decisions in Smart Cities
The economic value of IT is a well-established research stream within the information systems and management disciplines. While research traditionally put a stronger focus on the value to the private sector, i.e. the business value, more recent efforts have also emphasized the importance of considering the public value of IT. With the rising relevance of big data analytics applications, the value of this specific kind of IT investment is also receiving increased scrutiny, again with a focus on a business context. Clearly, though, the transformative and disruptive power of big data is not limited to the private sector. In this paper, we expand the state of knowledge on the value of IT by specifically considering the value of analytics in the public sector. Building upon Moore’s concept of public value, we outline differences between the public value of analytics and of other public sector IT applications, such as e-government. We position analytics within Moore’s strategic triangle of net value, legitimacy, and operational feasibility, proposing that analytics can provide public value through, for instance, economic benefits, societal benefits, and basing governmental decisions on objective criteria. We further investigate these propositions through an extensive demonstration case that leverages analytics capabilities to improve municipal decision-making. We focus on infrastructure planning, a crucial responsibility of public administrations that includes long-term strategic decisions with high initial costs and uncertain payoffs that materialize over several years or even decades. Our showcase concerns the issue of establishing a public charge point network for electric vehicles, a challenge cities around the world are currently facing. Using data from the city of Amsterdam, we introduce a spatial analytics application that combines conventional (demographics) and novel (points of interest) data sources to identify key determinants of charge point usage and, based on these determinants, prescribes a distribution of the charge point network that maximizes expected utilization for any city. Our results show that spatial analytics can improve charging point utilization by up to 15 percent. We derive several implications with respect to the public value of analytics. First, if charging demand is assumed to be fixed, necessary investments to satisfy this level of demand are reduced by 15 percent. For the case of Amsterdam, this translates into savings of about one million euro. Due to its applicability to any city, the potential savings resulting from this particular applications at a global scale would be even more substantial. Second, if an improved charging network encourages people to switch to electric mobility, a 15 percent increase translates to a reduction in CO2 emissions of 400 metric tons per year just in Amsterdam (if electricity is provided through renewables). Third, we outline the public value of analytics that derives from increased legitimacy of governmental decisions. Data-driven analytics provides objective indicators for strategic decisions, thereby increasing the transparency of the decision-making process. We conclude by discussing cost aspects and solution approaches through open data.