Julienne Molineaux writes of the risks in merging important civic archives.

Government agencies are often restructured in the belief that the change will lead to better performance.

But whether it is organisational design or information technology projects, bigger scale doesn't always mean better results. In fact, larger scale brings with it many problems.

Yet the Government and the State Services Commission are proposing a merger between Archives New Zealand, the National Library and the Department of Internal Affairs, with economies of scale in IT as the driver of the change.

The New Zealand state sector has a history of IT project failures, the $100 million-plus police INCIS endeavour being the most infamous.

The bigger the IT project, the bigger the concentration of risk.

The more complex the project, the more likely it is to flounder, go over-budget, go over-time, fail to deliver on its promises and be difficult for users to operate.

In proposing the merger between the archives, the library and Internal Affairs, State Services Minister Tony Ryall has admitted there are no burning problems that need solving. Rather, the rationale is the desire to develop an overarching platform from which New Zealanders can access civic information.

The merger is being promoted as providing a single decision-making centre in public technology services "to determine a whole of government approach to managing information".

While the current fragmentation does indeed make accessing information more difficult for the public, a serious question has to be raised. Just how feasible is the Government's dream of a unified civic information "super" platform?

If it is not feasible, the IT project at the base of this merger will end up being yet another expensive government technology failure.

Otago University academics Robin Gauld and Shaun Goldfinch describe the biggest threat to the success of government IT projects as a "dangerous enthusiasm", which causes managers to overstate the benefits of IT and blinkers them to potential problems.

Some degree of failure - whether in terms of timeliness, budget, specifications or usability - may be the norm in large-scale IT projects. But dangerous enthusiasm for such projects, Gauld and Goldfinch assert, makes failure even more likely. Good management oversight of IT projects is not sufficient to overcome the problems created by this enthusiasm.

The case for the merger of the three agencies reeks of dangerous enthusiasm. Official papers tout the IT outcomes of the merger in utopian terms. It will "[maximise] opportunities for New Zealanders to fully participate in government processes" and assist government to "exploit digital capability effectively to manage information and to ensure effective stewardship of and access to information held within the 'online' domain". This a very ambitious set of goals.

Presumably the new platform will initially focus on the digital records of the archives and the library. Even at this limited level, the project will not be straightforward. Archives and libraries have different classification systems, classify different types of materials and the two sets of professionals, archivists and librarians, have different work practices.

To be successful, any new system will need to integrate these differences in a way that keeps staff on board and recognises the logic behind their different approaches. Another alarm bell about this merger relates to the bigger scale and scope of Internal Affairs.

Official papers reveal that Internal Affairs' current hub and spoke structure is at capacity. Nonetheless, it will take on two new agencies and the role of co-ordinating the Government's new digitisation of civic information strategy.

The State Services Commission argues that this extra scale will bring efficiencies; but it could just as easily generate diseconomies arising from extra layers of management and reporting. State Services will reduce the number of chief executives it has to mentor from three to one - but that one chief executive will have a much more complex management task to get right.

The strategy, as outlined, has almost been set up to fail: it is large-scale, complex, faces resistance from staff, complicates the management task and involves a dangerous enthusiasm for IT.

The problem this merger purports to solve is described as not very pressing, but the likelihood of failure for its IT goals is exceedingly high.

And yet there are important digital information issues in the New Zealand state sector. A 2006 survey revealed that 77 per cent of responding government agencies had no process in place to preserve their digital information.

A recent survey by Archives New Zealand found that 53 per cent of agencies had digital information they could no longer access. The proposed merger focuses on front-end use of government information while ignoring serious back-end deficiencies.

There can be only one outcome of such an approach - gaping, irreversible holes in the content of the unified civic information platform.

Julienne Molineaux's PhD was on the history of government policy and machinery of government issues for the official archives in New Zealand, with a particular focus on the 1994-1999 period. She is a lecturer in the Centre for Business Interdisciplinary Studies at AUT University.