Lies, damned lies and the stocktake

There are, said Mark Twain, lies, damned lies and statistics. Never was the famous quote more appropriate than when applied to the government's information technology stocktake, published in November but conveniently released a week before Christmas when all the newspapers seeking the information had closed for the year. On the face of it, the statistics make quite impressive reading. Of 174 major projects (over $500,000) running or initiated during the past five years, just 14% were over budget, by an average 2%. Three had been dropped, three were on hold and three were 20% or more over budget But - and it's a big but - budget data was provided for only 80 of the 86 projects completed and for just 62 of the 82 in progress. In addition, 12 departments specified 44 projects that were not to be identified, either for reasons of commercial sensitivity or because of security considerations.

There are, said Mark Twain, lies, damned lies and statistics.

Never was the famous quote more appropriate than when applied to the government’s information technology stocktake, published in November but conveniently released a week before Christmas when all the newspapers seeking the information had closed for the year.

On the face of it, the statistics make quite impressive reading. Of 174 major projects (over $500,000) running or initiated during the past five years, just 14% were over budget, by an average 2%. Three had been dropped, three were on hold and three were 20% or more over budget.

But — and it’s a big but — budget data was provided for only 80 of the 86 projects completed and for just 62 of the 82 in progress. In addition, 12 departments specified 44 projects that were not to be identified, either for reasons of commercial sensitivity or because of security considerations.

You don’t have to be a Rhodes scholar to work out that it would take very few projects of those for which no budget data was provided to run badly over budget for the statistics to be well skewed.

The report also notes that researchers suspect that non-IT spending on IT would inflate formal budgets by at least 40 to 50%. “This seems to reflect that the IT budget is no longer an accurate picture of total IT expenditure.”

Of the 174 projects, 165 were described as mission-critical but risk assessment was undertaken for only 116. That leads the report to call for more emphasis on risk assessment.

Other figures of note:

• Expenditure on hardware and software during the past three years was $369.5 million.

• 1996 IT expenditure represented 0.35% of government revenue that year, half the international average.

• Major projects completed during the past five years cost $186.2 million — 2% under total budgets.

• During the past three years $277.1 million was spent on outsourcing.

The report lists a number of projects and their status (completed, on hold, in progress) but in no case attaches a budgetary figure. At some stage the complete data- set will be presented as a database on the Web at www.govt.nz.

In some respects the report damns itself under a section headed “The information base — understanding the data”. It says: “The stocktake did not require departments to verify their responses by, for example, providing relevant documents such as departmental accounts, policy statements, procedure manuals, and progress and audit reports. It is therefore not possible to state how accurate the cost figures are, whether procedures are actually used, how well applications work, or why a given project succeeds or fails.”

At best, the report may be described as indicative.

Join the newsletter!

Error: Please check your email address.
Show Comments

Market Place

[]