Association for Public Policy Analysis & Management

Despite the significant penetration of information technology (IT) innovations into organizations over the last two decades, we have limited understanding of how IT innovation adoption and implementation affect their performance (Buffat, 2015; Hood and Margetts, 2010; Kelman, 2007; Pang, et al., 2014; Pollitt, 2011; Zammuto et al., 2007). In this paper, I analyze the impact of the IT implementation on public sector performance. The IT implementation considered here is the 311 Customer Relationship Management (311/CRM) system. The 311 customer contact centers handle non-emergency service requests within local (i.e. city/ county) governments. The IT system tracks customer service requests from their initiation to their closure. The 311 centers are significant to study since they epitomize IT innovation adoption and implementation in the local government, effecting a potential sea-change in public service delivery. Presently, there are about 300 cities and counties with the 311 centers. About 30 of them use the standard Open 311 protocol for the service request data and are comparable across cities; hence 311 big data form the basis for this paper’s analysis. I focus on five departments which are known to receive a most service requests: Streets, Public Works, Waste Management, Parks and Recreation, and Non-emergency Police. The unit of analysis is thus the department.

To examine the relationship between IT innovation implementation and department performance over time, panel regressions with fixed effects at the department level is used. The fixed effects model eliminates all unobserved heterogeneity among organizations and controls for external factors that do not vary over time or are specific to a particular department (which addresses the omitted variables problem). This technique will allow to examine the performance effect of every additional year of 311/CRM (i.e. the IT implementation) in each department over time. We also do not have to be concerned about comparing different departments from different cities with each other (“apples and oranges” problem), since we only compare departments at point t+n with themselves at point t.

The performance (dependent variable) is measured as the average time to close a service request by a department (obtained from the 311 big data for the 30 centers). All variables will be based on data for every department i at time t. The independent variables are the number of years the 311 has been in operation at any particular time, which captures the innovation implementation (post innovation adoption by the organization). The beta values are estimable parameters, μ_i is the department-specific constant capturing time-invariant effects of unobserved factors, and ε_it is an error term.

Performance_it = β_{0i +} β₁311_it + β₂311_it² + μ_i + ε_it (1)

The quadratic component (311_it²) allows to examine non-linear relationship and whether there is a tipping point beyond which the effect of innovation on performance may rise or fall.

The overall contribution of the paper lies in ascertaining how the 311 CRM implementation affects organizational performance of the departments. The hypothesis, of course, is that the 311 implementation would be significant in enhancing the organizational performance of the departments.