If the four business outcomes can be achieved without OIMT having to take on the role of approver, OIMT would be much better off. Being an approver looks better on paper than in reality. There will always be pressure to take shortcuts and approve projects that don’t make sense in terms of the strategic plan (and sometimes don’t make sense at all). That’s because in the State "what the right things are" are often: 1. Get the money. 2. Spend the money. In state government it is a cardinal sin to turn down or turn back federal dollars. As long as the requirements of the grant are met everything else becomes secondary. As an approver OIMT will be right in the middle of this game and end up using 80% of its resources on 20% of projects that are nothing but a headache that someone is trying to get rid of. There’s a better way.
I suggest that the OIMT begin with a review not to approve a project but to determine whether or not it wants to become involved. A review for all IT purchases that exceed the amount required by HRS 103D for small purchases (don’t hard code the current number) would be a good place to start. This psychologically ties the review to existing processes so it seems like an extension of current rules and not a whole new set. By the time a program has decided to forgo the small purchase procedures and pursue a RFP or RFQ it's braced for the long haul. At worst the OIMT review will be seen as a minor inconvenience and may even be welcomed as helpful.
When the purpose of the review is to determine whether or not it makes sense for OIMT to apply its limited resources to the project, OIMT can stay more focused. There will be $200,000 COTS projects that are low risk, low priority in terms of the strategic plan, but high priority in terms of a program or a grant. There will also be $80,000 build from scratch projects that are high risk but give OIMT access to resources that push forward some strategic central element. Sometimes OIMT will just have a full plate and can’t take on another project without reducing the overall quality of its work, and yet, a program won’t be able to wait and will need to move forward now to meet grant requirements. OIMT needs to implement a safety valve, the ability to return the project to the program with no further review requirements.
Is OIMT relinquishing the control of IT resources in these cases? A lot of money coming into the State is earmarked. In these cases OIMT will not have the flexibility to divert the money to more worthwhile projects. So in reality OIMT had no control over the money to begin with. It needs to decide whether or not to fight the program in order to take control. In most cases it just won’t be worth it, the energy would be better expended elsewhere. OIMT needs to retain the right to make these types of judgments until it has unlimited resources. Then it can take responsibility for approving everything.
The same could be said for the ELC. Let them do a preliminary review and decide for themselves the projects they want to look at in detail. This is the model the Supreme Court uses. If the Supreme Court had to hear every case it would quickly become overwhelmed.
At the small purchase level OIMT needs to question why it’s getting involved at all. There’s not that much to gain, and in the big picture any apparent efficiency gained on the OIMT side is going to be paid for on the program side. Programs waste a lot of resources getting permissions for inconsequential purchases from people that know nothing about the program. Is it possible to enlighten approvers to the extent that they can make as good a judgment as a branch chief? Is it worth the effort? Not in the case of small purchases. I spend so much time justifying forgone approvals that I often feel like my primary job is creative writing, not IT. The goal is to please my audience.
Instead of approvals, pursue standardization at the small purchase level through the OneNet model. Become the vendor of first choice. In the meantime, control standards by getting DAGS to issue price lists instead of vendor lists. Keep WSCA but use lists of approved items to change WSCA from a superstore to a supply depot. Only require review of exceptions. Exceptions will almost always be part of a larger project that will have been reviewed anyway. Drawing attention to exceptions is also a good way to catch large projects that haven’t been submitted for review.
Make it a priority to get DAGS to come up with a statewide leasing contract for desktops. Actions like these will go a long way toward standardizing purchases of commodity items so that the State can provide central support for them, AND, at the same time eliminate any need for small purchase approvals for technical reasons.
If it can, OIMT should convince DAGS to get rid of the T-205 altogether. If it can’t OIMT should divorce itself from it completely. The T-205 has a bad reputation and a bad image. If a department wants more control over small purchases they always have the ability to implement stricter standards. Let the department be responsible for obsolete procedures, not OIMT.
Acknowledging up front that OIMT cannot be omnipotent in the State’s IT world and letting programs have final say on some decisions will improve OIMT’s image, but not enough. The governance plan reflects the org chart too well. It does not take into consideration the org chart’s limitations. Directors come and go. Branch chiefs are forever. The primary reason the civil service system was created was to provide continuity of government services regardless of changes in administration. In order to do that the civil service system does in fact place limitations on the power of an administration. You can see why this is necessary. A new administration could come in, replace all the managers with friends and relatives, and then proceed to blow most of the budget on pet projects, not leaving enough money to provide adequate essential services.
In spite of what the org chart says, directors and program managers share power. In the case of earmarked funds, program managers have the lion’s share of the say. This is a very big way that government differs from commercial enterprises. This plan seems to forget that. It gives Director’s the same amount of authority they would have in an enterprise, much more authority than they currently have now. OIMT then proceeds to assume much of that power for itself. It is really bad public relations. It makes it look like an attempt at a power grab.
OIMT would do better positioning itself as a service provider and educator, a kind of internal consultant. OIMT would have a lot of influence and avoid the political wars.
Program managers absolutely hate it when they think someone is telling them how to run their program. They best understand the limitations imposed on them by the laws covering their program. They rightly believe that they are in the best position to make business decisions, not some upstart Director or IT group. On the other hand program managers love it when someone really helps. The program managers are happy to pay for that service (except for having to get payments through the approval process). And that brings me to my last suggestion.
OIMT should investigate becoming self-supporting along the lines of the special fund model. Besides earmarks, another big reason for the variability in IT maturity is that general funded programs are always maintained at subsistence levels and even those levels are always being tested. Nobody wants to be seen as responsible for the death of a program. By keeping general funding for any given program at an absolute minimum, there’s more money to spread around and the total number of programs in existence can increase. Everyone can take credit for supporting programs without having to take blame for eliminating others. Federally funded programs get just enough State funding to make match. It’s the earmarked federal money, which the program manager controls, that gives them the ability to invest in IT.
Using a fee for service model has several advantages. Instead of asking for big-ticket items, it makes it possible for OIMT to spread out costs by having programs put a barely noticable line item in their budget for information technology services. It also gives OIMT an easy way to draw off earmarked funds. It’s not a perfect solution. I don’t know if it’s ever been done for internal services. It will take a legislative fight and even if OIMT gets special fund status the legislature can always raid the fund, but it beats being in a position of constantly just scraping by.
DAGS uses a fee for service model to charge special and federally funded programs for accounting services. In their case however it doesn’t work. The money ends up in the general fund and is then given to other programs that have more political visibility than an accounting program. DAGS ends up providing lousy service due to a lack of resources. In some cases this happens directly and in other cases it’s indirect but it’s effectively the same result. The money goes to other services and people don’t get what they paid for.
I’ve talked to several people with more than 20 years experience in the State and they all think the OIMT plans are too ambitious and that they will never be implemented. I agree. It’s just too much to digest all at once. I can’t imagine a legislator taking the time to look at these plans in detail. On the surface it looks like an IT proposal. In terms of political visibility, IT falls much closer to the accounting side of the scale than the education side. But all is not lost, a combination of service orientation, support from the programs, and fee for service could give OIMT a chance to bootstrap these plans into existence when the business argument fails.