Often, IT companies sell e-government to politicians, and politicians sell e-government to citizens on the promise that it will save money. These claims regularly appear as “news” items, especially in IT- or government-related media. This has in part encouraged the huge expenditure on e-government: a ballpark figure would be US$3 trillion during the first decade of the 2000s[1].
So here’s my question: “If e-government is so great at cutting costs, how come my taxes haven’t gone down?”
Of course, taxes depend on far more variables than just e-government. But the simple answer to the question is “. . . because e-government does not save money, it costs money”. That seems likely the case in the global North where e-government seeks to cut costs by replacing expensive humans with cheap technology. It is most definitely going to be the case in the global South where the technology is more expensive and the humans are much cheaper.
Despite the obvious importance of the topic, good quality cost:benefit calculations are rare but can be found. Six years of e-government in UK local government saw £3.90 billion of investment release just £0.97 billion of savings[2]. The aggregate cost:benefit ratio of e-government projects in Australia was 1.64:1[3].
Rarer still is good quality work from developing countries.[4] However, a recently-published study of e-government in Bhutan by Mayumi Miyata provides a model for systematic and comprehensive evaluation of e-government costs and benefits. The case study focuses on the Road Safety and Transport Authority of Bhutan, which issues driving licences and vehicle registration documents. This was traditionally a paper-based process, and often slow; particularly for driving licences which had to be sent by post from regional offices to the head office in Thimphu. In the mid-2000s, an Internet-enabled database system was installed so the main information associated with these processes could be passed instantly between offices. (This was therefore an “e-administration” application for use by government staff rather than an “e-services” application for use by citizens.)
Data for Miyata’s research was gathered both before and after the introduction of this e-government system including detailed observation and timing of work processes, a breakdown of departmental accounts, and a survey of citizens. The “after” component was undertaken in 2007; two years after implementation of the system, allowing plenty of bedding-in time.
Activity-based costing showed that the direct labour cost for issuing licences and registrations fell 24% following introduction of e-government; from US$15,080 to US$11,530 per year[5]. For example, the direct cost of issuing one driving licence fell from US$1.57 to US$1.17. This was achieved largely through a significant redesign and decentralisation of internal decision-making and workflow.
However, introduction of e-government brought additional costs – hardware, software, internet connectivity and the cost of IT staff – totalling US$11,080 per year (set-up costs being amortised over 10 years). The only indirect saving was in reduced postal cost (US$720). Thus, overall costs were US$15,800 per year before e-government; US$22,610 after e-government. A rise of 43%.
We need to recognise some specific features of this case that make it typical of a least developed country:
- the particularly low labour costs and high IT costs;
- the relatively low volumes of transactions across which costs can be spread (the case is more akin to a local than national government in size);
- the use of e-administration rather than a web-based self-service system which, while still requiring human back-office intervention, would automate some processes.
Miyata’s research thus provides a model that should be replicated for a broader set of examples.
On the other hand, Miyata’s work misses out three additional reasons why e-government globally fails to deliver cost savings:
- the relatively high rate of e-gov project failure, the costs of which must be included in any overall cost:benefit accounting[6];
- the learning curve – often of some years – that must be traversed before e-government applications can be used efficiently and effectively[7];
- the need for government e-services to be run in parallel with existing face-to-face, phone and postal service channels in order to bridge the digital divide and avoid excluding large sections of the population from access to government services; public e-services thus being a supplement to, not substitute for, other channels[8].
Does this mean e-government is a waste of money, and we should ask for our US$3 trillion back? No. What it means is that e-government is not going to save money for government and help bring taxes down. The benefits of e-government lie elsewhere. Again, Miyata’s paper is a good illustration:
- External savings: the lead time from application to receipt was reduced by minutes, weeks, even months for outlying offices. Wait times in offices may also have come down. Other studies report shorter waits and fewer journeys. Saving of journeys can be monetised, and saving of citizen time might be (it depends how that saving is spent). The key cost savings of e-government may thus be external not internal: for service users not administrators.
- Internal control and accountability: e-government provided managers with greater oversight of work processes and staff.
- Service quality and equity: citizens reported the quality of service and the fairness of treatment improved after introduction of e-government.
Other research shows further qualitative and external benefits delivered by e-government including: greater transparency of public services; greater accountability of public servants and politicians; reduced corruption; lower costs for business; greater attraction of foreign investment[9]. Please comment to add your own examples of evidence.
So e-government may not bring your taxes down, but – if properly designed and implemented – it will bring a positive economic and social return on investment.
[1] Heeks, R.B. (2006) Managing and Implementing eGovernment, Sage, London http://books.google.com/books?id=hRzAnMulatUC&dq; WITSA (2008) Digital Planet 2008, World IT Services Association, Kuala Lumpur, Malaysia; see: http://www.witsa.org/KL08/DigitalPlanet2008_ReportTables.pdf
[2] Kable (2005) Implementing Electronic Government 4, Kable, London
[3] Foley, P. & Ghani, S. (2007) The Business Case for e-Government, paper prepared for High-Level Seminar on Measuring and Evaluating E-Government, Dubai, 12-13 March http://www.oecd.org/dataoecd/44/42/38404094.pdf
[4] There is a good study of e-government projects in India but it was unable to capture cost data, so focuses only on benefits.
[5] These are costs for issuing just over 31,000 documents. Note this excludes the cost of materials for licences/registrations, which was the same before and after e-government.
[6] Heeks, R.B. (2006) Managing and Implementing eGovernment, Sage, London http://books.google.com/books?id=hRzAnMulatUC&dq; Gauld, R. & Goldfinch, S. (2006) Dangerous Enthusiasms: E-Government, Computer Failure and Information System Development, University of Otago Press, Dunedin, New Zealand
[7] Poostchi, M. (2003) Implementing E-government, MBA dissertation, Carleton University, Ottawa, ON
[8] Helbig, N., Gil-Garcia, J.R. & Ferro, E. (2009) Understanding the complexity of electronic government, Government Information Quarterly, 26(1), 89-97
[9] Accenture (2004) eGovernment Leadership: High Performance, Maximum Value, Accenture, Dublin; Bhatnagar, S. & Singh, N. (2010) Assessing the impact of e-government: a study of projects in India, Information Technologies and International Development, 6(2), 109-127 http://itidjournal.org/itid/article/viewFile/523/231
Follow @CDIManchester
ICTs for Development 
I feel the analysis of a small, isolated e-application does not represent a fair evaluation of global potential benefits of e-gov. Actually, I never heard of the presumption that e-gov would reduce your personal taxes. On the other hand, there seems to be some evidence that well applied e-gov WILL reduce waste in IT investments, particularly if a SOA paradigm has been applied, will reduce consumables expenditures (scale economy) and will increase tax collection by closing tax evasion and tax elusion cracks.
Now, what a government chooses to do with those budget gains is a political issue, not that of e-gov in itself. Offering more services and employment maybe more interesting than reducing your taxes by a pound or two.
Also, if the study mentionned does point at valid cons, particularly for “southern” countries such as replacing cheap labor with expensive technology, a long term analysis of present and future value of the investment might turn out another result.
And lastly, “southern” countries have often a different set of financial requirements than the “northern” ones, particulary with respect to access to external financing being actually easier to find, for IT projects, than funding local, recurrent costs within budgets that are for ever unbalanced. Foreign currency for investment is easier to find than local currency for recurring costs.
However, Nicaragua offers one interesting case study (the Registry of Governmental Suppliers) where a mixture of not too expensive technology with low cost labor was intelligently applied and which might deserve a study: by creating “help offices” accesible on line, some parts of the administrative processes can be executed manually as back-office, while allowing users to access the processes on line with a very acceptable 24 hours reply, without having to be physically present. For occasional processes, a 24 hours wait without leaving the house or the office can be quite satisfactory.
Agreed on all the main points – just to be fair to the author (!) on your first point, the Bhutan case is an illustration of wider trends that are evidenced in the sources listed in the footnotes; so this is not a crazy extrapolation from that one case.