Complex efficiency benchmarking of public agencies applied on Estonian rescue brigades

Improving the management of public agencies towards better use of inputs is a difficult task, both theoretically and practically. Theoretically, the outcomes of the work of public agencies need to be defined as well as the utility of the direct outputs for the various stakeholders. The optimal size of the potential (readiness to offer the service) of the public agencies also needs to be derived from the direct outputs. The basic elements of the work of public agencies (inputs, potential, outputs, outcomes) form a complete system of performance which can be characterized by the ratio of outcomes to outputs, potential and inputs (effectiveness), the ratio of outputs to potential and inputs (output efficiency), and the ratio of potential to inputs (process efficiency). Performance is therefore a complex phenomenon, which must be comprehensively described using a system of indicators developed using appropriate modelling methods.

In the present study on the base of the benchmarking approach, the frontier analysis methods (data envelopment analysis and stochastic frontier analysis) are used. To evaluate the link between socio-economic environmental factors and the efficiency estimates a regression analysis is used. A deterministic synthesis method is used to obtain an integrated assessment of the efficiency of public agencies.

The research focuses on 66 national rescue service brigades in Estonia as the regional subunits of the Estonian Rescue Board over 2011–2012. The total amount of allocated budget and machinery park size are used as input indicators for the brigades. Potential is described using speed of response in the case of a rescue event and the number of residents reached within a certain timeframe (e.g. 5, 15 and 30 minutes). Outputs are described according to the amount of various rescue events and rescued people. The socio-economic environment is described according to service area, population size and density, number of hazardous objects and car and rail traffic frequency in the region.

To benchmark the efficiency estimates for the rescue brigades, two simplified and two multi-component models are constructed using the DEA and SFA methods. Different model types reflect the efficiency of the rescue brigades from different perspectives and the calculated estimates (as well as rankings) are only weakly correlated with each other. According to the results, we conclude that the frontier analysis methods allow us to identify "best efficiency practices", which take into account the covariance of the different elements of the performance of public agencies.