The Article makes a key point. “Modernization” (rearchitecting, rewriting, retesting) of these major back-end applications is expensive and takes time. Lots of money, too much to be committed by a government at one allocation. Too much time to be done in one administration. And, like most really big projects, estimates of cost and time are rarely large enough.

We have heard of firms with huge legacy application code bases committing over a billion dollars recently. That kind of money is almost impossible for a State Government to marshal. Even the US Government would find it hard to allocate so much to such a mundane sounding project. Sure, the cost of running those mainframes is higher than seems reasonable in a world of thousand dollar laptops and The Cloud. But spending that kind of money, even if you are successful, gets you a new application that does, at least, what the old one did. We find it hard to understand how a project like that meets normal justification criteria.

And then we find out the New Jersey problem was probably a swamped Web front-end. Ironic.