eGovernment for Development

Assessing eGov Project Risks:
Real-World Design-Reality Example 2

Computerising Election Results Management in West Africa

• Background
• Risk Assessment
• Risk Mitigation

Case Study Authors

Kwame Boateng and Richard Heeks

Background

The National Election Commission (NEC) has 1,200 staff and a four-layer structure (national HQ, regions, districts and constituencies). It has been the target of vehement complaints from opposition politicians, and the subject of waning confidence from the electorate, following a set of elections that are seen as having been poorly managed. Driven by a need for both actual and perceived improvement, a computer networking project has been designed that will transmit results direct from the 200 constituency offices to the national results declaration centre in the capital. This will bypass both the current distrct and regional NEC structure and parts of the national structure, all of which has been viewed with suspicion by some politicians and citizens. As well as improving the transparency of the election process (which will hopefully reduce the incidence of post-election violence), there will also be an increase in speed of results delivery. As well as NEC staff, all citizens and political parties are stakeholders in the process. Foreign governments and donors are secondary stakeholders.

Risk Assessment Via Design-Reality Gap Analysis

Design-reality gap analysis compares the assumptions/requirements within the application design with the existing reality in the public organisation along seven 'ITPOSMO' dimensions. The larger the gap, the larger the risk for the e-government project. Follow this link for more information on design-reality gap analysis.

• Information : the design does not aim to significantly alter the reality of the content of electoral result information except that information which is falsely returned. It will alter the steps of information flow but not the overall start and end points. Design-reality gap rating: 4.5.
• Technology : the design assumes an electronic scoreboard at national headquarters plus c.350 networked PCs, with one in each constituency office. The current reality is no computers in any constituency offices and c.50 in regional and national offices of the NEC. Most constituency offices and all at higher levels have faxes. Twenty constituency offices have no electricity or phone connection, and 25 have an inadequate building structure. Design-reality gap rating: 7.5.
• Processes : the design alters the current results process through electronic disintermediation. The current reality is that constituency results are faxed to regional NEC directors, who then fax those results to the NEC Director (the only person legally empowered to declare any election result), who then hand-writes the results on a national scoreboard. Under the new design, constituency returning officers are empowered to declare results once those results have been signed by all candidates' agents. Those results will then be keyed into the constituency PC, with the result appearing directly on the national electronic scoreboard. The changes only relate to results management: voters themselves will still vote manually, and those votes will still be hand counted. Results still flow from constituency to capital. Design-reality gap rating: 5.
• Objectives and values : the design assumes an objective of transparent reporting of election results by an accountable NEC that values free and fair elections. In reality, particularly given the former opposition party's election to power in the most recent elections, most stakeholders share these objectives and values. Senior NEC staff also appear to support these design assumptions, recognising how low their current standing is with citizens and politicians. There are just a few hints of resistance, e.g. from district staff who do not share the system objectives. Design-reality gap rating: 3.
• Staffing and skills : the design assumes a range of competencies to be present. These relate partly to the installation of the network and other hardware/software, but particularly to the need for data entry skills and network operation/maintenance skills at election time. In reality, the NEC's IT unit has only half the required installation/maintenance staff, and there are no staff currently employed by the NEC for data entry work in constituencies. There is a small gap related to returning officers learning the new results-handling process. Design-reality gap rating: 6.
• Management systems and structures : a design decision was made to retain the existing hierarchical structures of the NEC. Although regional and district offices will no longer handle the returning of results, they retain all of their traditional 'in reality' roles of administering the electoral process. Design-reality gap rating: 2.5.
• Other resources : the design assumes a total cost of US$20m: US$3m from the national government for upgrading constituency offices; US$11m from one donor agency for constituency PCs and network links; US$6m from another donor for central hardware and software plus training. The design thus matches the reality of both actual costs and money available. The design also assumes a sub-two-year timescale for the project, enabling full computerisation in time for the 2004 local elections. In reality, this completely mismatches the actual time available. Design-reality gap rating: 5.5.

Overall Risk Assessment and Recommendation

The overall gap rating total for this design proposal is 34. According to the gap assessment table, this suggests a possibility of partial or even total failure of the project unless action is taken. (Note: at the time of writing, the project was already heading for failure because of its over-optimistic time schedule. At a point when 90% of the IT should have been in place, only a few computers had been delivered, and work on constituency offices had yet to occur.) The recommendation was to proceed with the project, but undertaking actions to reduce some of the larger design-reality gaps.

Gap Reduction Actions to Mitigate Project Risks

To mitigate e-government project risks, you first identify those ITPOSMO dimensions with the largest gaps. Then you attempt to reduce those gaps by altering the project design to make it more like current reality and/or by altering current reality to make it more like the project design.

The three largest gaps occur on the technology, staffing and skills, and other resources dimensions. It is these dimensions that should therefore be addressed first.

The proposed risk mitigation actions here were:

1. Other resources/Technology dimensions . Alter project design by increasing the timescale for the project; recognising that full results computerisation for the 2004 local elections is not feasible, and that the project should aim instead for a pilot in 2004 in one region, and for full results computerisation by the 2006 national elections. This would have the beneficial knock-on of stretching the technology (and other dimensions) gap over a longer time-frame, thus reduced the size of design-reality gap in any given time period and, hence, reducing project risk. Piloting in 2004 will avoid dangers of a growing objectives/values gap which would happen if citizens, politicians or donors lose their momentum for the project.
2. Staffing and skills/Technology dimensions . Alter current reality by increasing the competencies of NEC staff. In part, this will mean identification of available data entry staff for the election period. This will also mean IT- and process-related training for permanent staff members. Uses for the constituency and other PCs should be found other than just returning results. This will ensure a gradual integration of the technology and related skills into the work of the NEC, thus helping a gradual reduction of these dimensional gaps.
3. Staffing and skills dimension . Alter current reality by hiring an additional systems analyst/designer, and two additional network installation/maintenance staff for the NEC's IT unit.