Alberto Savoldelli, Gianluca Misuraca and Cristiano Codagnone
Government projects, with a more transparent and participatory assessment process since the early stage of the policy decisions( Savoldelli, Codagnone & Misuraca, 2013).
Figure 2: Key drivers of e‐Government adoption( source: Savoldelli, Codagnone & Misuraca, 2013)
3. Key dimensions of eGEP‐2.0 as an impact assessment model of policy decisions
The simplest method to represent the consequential chain of events to be considered for an impact assessment is an Analytic Hierarchy Process( AHP) model( Saaty, 2008). From a policy‐model perspective, an AHP starts from the ultimate policy goal and it is organised in several levels linked each other with mutually dependent relationships. In literature several examples of AHP applied to policy decision‐making process exist( Mansar, 2006; Kahraman, Demirel, & Demirel, 2007; Parra‐Lopez, Groot, Carmona‐Torres & Rossing, 2009; Saaty, & Vargas, 2012), however most of them adopt AHP for modelling projects ' selection and / or evaluation purposes, and few of them has been used for modelling policy‐making decisions. Also eGEP( Codagnone et al., 2006), that is our starting point in the development of eGEP‐2.0, does the same. eGEP is structured into three levels: overall goal( level 0); outcomes criteria( Level 1); impacts indicators( Level 2), and it has been applied for benchmarking the performances of e‐Government web portals across administrations belonging to different EU Member States, therefore its nature mainly descriptive was successfully used in ex‐post assessment of e‐Government initiatives( Codagnone & Undehim, 2008). However eGEP as it was designed cannot be used to model a policy decision‐making process as eGEP‐2.0 aims to do, mainly because:
• it does not allow to represent the policy strategy formulation process, from the definition of the overall goals of a given policy planning cycle, to policy guidelines definition and approval( Savoldelli, Codagnone & Misuraca, 2013);
• it does not provide the necessary decisional links between the policy strategy dimension and the implementation projects dimension( Lundqist, 2006; Suggett, 2011);
• it does not have the ability to model the circular nature of the policy‐making process( see figure 3 and also Heeks, 2006; Savoldelli, Codagnone & Misuraca, 2012).
In general terms, in fact, expected policy outcomes shape expected policy impacts( usually called political guidelines). For achieving these guidelines, public administrations implement projects which start their lifecycle as project ideas, by defining projects ' expected outcomes, usually called needs. Upon these needs, an ex‐ante estimation of the degree of achievement of quantifiable objectives is provided before deciding the projects ' portfolio suitable for implementing a given policy plan. These objectives are usually called expected impacts and are used for justifying project expected outputs. Vice versa the degree of achievement of given projects ' outputs, allows estimating projects ' impacts and to determine to what extent projects have satisfied the need for which they have been implemented.
Of course this is a simplification of reality, where there are no deterministic cause‐effect links amongst policy decisions – projects ' outputs – projects / policy impacts. As anticipated, only a cause‐effect analysis based upon a robust counterfactual approach can evaluate the probabilistic and stochastic relationships that more likely
445