The contract the ledger never saw
Here is a standard timetable for a contract. The contract gets signed on a Wednesday. Over three years and four milestones your office holds back five per cent retention from every payment, and the contractor had to post a performance security bond even before any work at all had been performed. Finally eleven weeks later the first invoice arrives at the accounts payable desk, and it is the first time the finance/accounting system (FMIS) has seen or heard anything about it. Just one number.
Now, this can be completely correct. Procurement did its part. Now it is the work of accounts payable. The normal situation, the typical arrangement from almost everywhere I have worked. Nobody hid anything or violated rules. An auditor would find that everyone followed procedures.
The big action by government is awarding the contract, done by procurement. Months later finance records an invoice, which often has nothing but a paper backup to connect the two. One office, procurement, runs the tender and all the related steps. Finance or Treasury pays the invoice. The contract, what ties the two together, is typically adrift between the two. A specialised “contract module” could connect the two, but was usually passed over due to budget constraints. (When the question is “can we still operate without it?” and the answer is yes, the follow-on question of “will this still function well” is almost never asked). The end result: no commitment record.
The consequence is not just a matter of counting beans. If the official commitment document is only the invoice, then when the contract is signed there is no obligation, and the obligation creeps up as payments are actually made. Not until you have to make the final payment is it obvious how much you have to make: if the contract was recorded on day one, the record would be clear from day one. This means that the biggest obligations, those large contracts, are the ones that the financial/accounting system knows the least about. Contrast this with the small immediate bills, where the full impact is recognized and funnelled through immediately.
What can make it much worse is that my example is only that of a simple contract. What if there are variations that add scope, add value or perhaps reduce it, stretch or reduce the time to finish the project? It becomes next to impossible to track or realise why change happened, and to hold those accountable. Or at least understand it so perhaps you will not repeat it. Few organisations have Public Investment Management (PIM) systems that track all of this in detail. Simply tracking the contract itself can feel like an Olympic moment. No one is trying to pull a trick, no one has to be deliberately deceptive, yet the end result could be the same.
After the signing, events and facts can pile up in finance that no one looks at. Retention can be costly, a cost vendors have to absorb (or raise their prices on). Performance securities are very important, but if the project stretches they can expire and become useless (good for a contractor, but then expensive if they have to renew it, unnecessarily). Liquidated damages are a real remedy, but not one to be cavalier about. Certifying milestones is an essential step. Each of these is indeed a fact which a system can track (facts are trackable). Typically, these are sitting in a contract file, in a folder that the accountant never sees (perhaps down the corridor or in another building).
Readers from last week will see that this is last week’s problem from the opposite side. Commitment control is one upfront question (do we go with the item). But a contract, especially for a project, often commits, then commits again, then even more times as modifications are pursued. That initial check for the commitment does not cover all those follow-up modifications, meaning problems could grow and grow.
Resolve this by realising that the first step, commitment control, does not let you just sit and rest, though that signature is of vital importance (necessary but not sufficient). Record the contract from the start when you get that signature, that official commitment, and track when invoices are presented, when modifications are made. Procurement’s initial award is an important financial event as well as a procurement event, and that financial event has to be recorded in finance, as alien as that function might be. This probably takes someone senior to take a position and make a stand, but unless this is done the initial procurement event stays isolated.
All these variations on a contract are in effect new contracts. The official approval, that commitment with the signature assigning approval and responsibility, needs to be provided each and every time. Otherwise, after the fact, finger pointing as to how the contract crept up. Everyone might tell you just to approve, approve, it must get done, but if you do not hold the line, in the end you have nothing but someone asking “How did the contract get so high? Who approved these changes?”
A solid change, a workflow or structural change, is to put the financial terms where the accountants can get them, not simply where the procurement officer sits. If the information is not easy to get, it will not be incorporated. If not incorporated, it will not be followed. Rather than make a complex system to fix this, to somehow change procedures, just take the simple and straightforward action of passing along the information. Something I would request for years, though you do have to get people to follow your requests.
Technology can route this. Put both procurement and accounting in one system. The difficulty is often more organisational or territorial, as procurement and the accounting (or treasury) office are different sections, and often only casually interact. Their jobs are different, they report to different people. Pass a baton in an open area, it can fall into the crevice.
This is not really conceptually that difficult. What is important is to recognise that the two functions, procurement and accounting, have to connect without dropping that baton.


