As a former chief information officer for the Obama administration, and former CIO for Microsoft, The Walt Disney Co. and VMware, I’ve learned a few things over the past four decades about early indicators of a technology project’s chances of success or failure.
Based on what I have seen to date, it is clear to me that the “free” tax-filing software project (so-called “IRS Direct File”) undertaken by the Department of the Treasury and the Internal Revenue Service is virtually guaranteed to fail.
While the reasons that a technology project succeeds or fails may vary, I’ve found that a basic set of questions must be addressed at the outset of any project:
Is there a compelling need?
In the case of IRS Direct File — which would have the government compete with private-sector tax software companies in the preparation and filing of American taxpayers’ individual federal tax returns — the government has conducted a few surveys over the years, and more recently has been urged on by a small group of members of Congress.
But such surveys do not translate into a reliable demand signal. Actual experience is a much more dependable indicator of whether there is a real and compelling need. And there’s an excellent recent independent analysis by the MITRE Corporation that comes to the opposite conclusion of these government surveys, is based on actual experience, and identifies the many flaws in those survey methodologies.
Thus, a clear, demonstrable need for IRS Direct File appears to be missing.
Even if there is a need for the technology project, is the proposed approach adequate and appropriate to address that need?
In this case, the proposed solution — tax prep software developed by the IRS for what seems to be federal tax returns only — will come up short, because it doesn’t address some of the most basic needs of the targeted taxpayers.
Among those needs are:
These are important matters that the commercial tax prep companies have long ago learned and are missing from the proposed IRS approach on Direct File.
Are the estimated total lifecycle costs — e.g., initial rollout, operational support costs, ongoing upgrade and functionality update costs, and the cost of additional cybersecurity — commensurate with the benefits?
Neither the IRS nor members of Congress who champion this IRS-sponsored approach have been able to characterize the benefits of their strategy in terms of time saved, costs avoided, increased customer satisfaction or any other commonly used metric of success/benefit. Nor have they been able to identify the expected cost per return (derived by aggregating all costs associated with the project and dividing that figure by the number of users).
The estimated costs of this project are very likely grossly understated, based on a sampling of other government software development projects. This is a pattern I’ve seen over and over, and it is not likely to change.
As the former federal CIO, I have a lot of respect for the men and women who are tasked with trying to implement such technology projects. But asking a very precious resource (IRS technical staff) to take on a project that gives every indication of being unnecessary, ill-conceived and doomed to fail is a waste of their time — and of taxpayers’ money.
There are many other high-priority technology issues that the IRS can and should focus on — not the least of which is upgrading its core infrastructure, which is underfunded and behind schedule. Here, with IRS Direct File, they are being asked to implement a project that, even if completed, will be extremely costly to operate and maintain, and likely would serve as a half-baked and little-used skeleton of already available and fully functional, market-tested commercial software.
Tony Scott served as the Federal Chief Information Officer in the Obama Administration.
