Public sector Budgeting and accountability Commitments and compliance

Custom public financial management software

Public money is allocated money. A representative body or board adopts a budget, which sets a ceiling per programme within which spending is allowed. Spending outside that decision is not a minor setback; it is a question of authority that resurfaces in the annual accounts. Appfront builds custom software for that chain: drawing up and amending the budget, recording commitments at the moment of award, cash and liquidity management, and a trail from annual accounts line to source document that an auditor can follow.

Public money works differently from business money

In a company, a budget is an expectation. Management sets it, and if revenue keeps pace the department may spend more than planned, because profit is the real criterion. In public finance, a budget is a decision. The amount per programme has been authorised by a council, a board or a representative body, and that amount is both the ceiling and the mandate. Spending more first requires an amendment to that decision, and only then a payment. This is called budget authority, and it is why public financial management software looks different from financial software for a company: the key question is not how much is earned, but whether every euro was spent within a valid decision and whether that can be demonstrated afterwards.

There is a second difference: the moment money is committed. A company that tracks its costs by invoice date has a workable picture. A public organisation does not, because a whole budget year can pass between awarding a contract and the first invoice. As long as that award is recorded nowhere, the budget looks free when it has already been spent. The commitments ledger is therefore not an accounting side issue but the core of the system, and the reason budget holders in the public sector work with three figures instead of two: budgeted, committed and actual.

This applies wherever money with a public purpose is managed, not only at a municipality or a province. Water boards, implementation agencies, joint arrangements and partnerships have the same questions, often in a harder form, because they account to several participants who each have their own budget cycle. Institutions in education, healthcare and culture that are largely publicly funded follow the same logic: they have their own operating account, but for earmarked funds they must account separately per scheme. And public funds and programme offices see the problem from both sides, as they are both recipient and provider. What binds all these organisations is that the criterion is not return, but whether the money was spent on the purpose for which it was made available.

Are you in the right place? This page is about public money: budget discipline, commitments in advance and accountability afterwards. If you are looking for analysis of business figures, such as margin per product or customer, variance analysis on results and group consolidation, that is financial analysis in a company and not this subject. If your question is mainly about reconciling bank transactions, subledgers and the general ledger, you belong with financial reconciliation software. Here the central question is whether an expense fitted within an approved decision.

The budget is a ceiling

The approved amount per programme is the upper limit. The system should show what is still available, not just what has been booked.

The commitment comes before the invoice

At the moment of award, money is tied up even though nothing has been paid yet. Without that registration, the budget holder steers on an outdated picture.

Accountability is the end product

The annual accounts are not an internal report but the document on which a board is judged and on which the auditor issues an opinion.

Building the budget, and keeping it accurate afterwards

A public budget takes shape in rounds. There is a starting point with the existing multi-year series, a set of assumptions on wage and price developments, and a number of proposals for new policy that are not yet in that series. Departments submit their input, the finance function adds it up, and then the cutting and shifting begins until the whole balances. In most organisations that process consists of a series of spreadsheets that go round by email each round, with the result that at the end nobody can reconstruct exactly which version led to the final decision.

Software only helps here if it models the process itself and doesn't just store the end result. That means: a budget round as an object with an owner and a status, proposals that exist as separate amendments with supporting rationale beneath them, and a scenario in which a board can see what happens if a proposal is dropped or postponed. Multi-year planning is not an extra here: public budgets routinely look several years ahead, and a proposal that costs almost nothing in the first year can upset the whole series in the last. Anyone who only sees the first year is making decisions with half the information.

Who may do what is already a question when the budget is being built. A budget is approved at a high level but executed by budget holders who each manage a part, sometimes with a further breakdown into sub-budget holders per team or project. That allocation is not an administrative detail, because it determines who may enter into a commitment, up to what amount, and who may approve an invoice. In many organisations that allocation lives in a decision that sits outside the system, so the authorisations in the software and the formal mandate slowly drift apart. Software that treats budget holdership as a given, with a validity period and a designated deputy for absences, keeps the two together and makes the after-the-fact review considerably easier.

Investments also need their own treatment alongside the operating budget. A credit is often made available in one go for a project that runs over several years, while the budget is approved per year. You are therefore monitoring two ceilings at once: the total credit over the project's lifetime, and the part that may be spent in the current year. On top of that, part of the project costs is capitalised and returns in later years as depreciation in the operating budget, while another part hits as an expense immediately. A system that treats credits as an ordinary budget line therefore almost always gives half a picture: it shows how much of the annual amount has been used without the position of the credit, or exactly the reverse.

At the same time, the structure of the budget must match the structure in which accountability is reported later. For Dutch local and regional government, the Besluit begroting en verantwoording (Budget and Accountability Decree) prescribes a fixed format: the budget and the annual accounts follow the same structure of programmes and task areas, with mandatory explanatory notes and a number of fixed sections on topics such as operational management, resilience capacity and maintenance of capital assets. Internationally, the IPSAS standards play a comparable role for public-sector reporting. We do not cite article numbers here, because those shift; what remains is the requirement that the structure of your budget, your general ledger and your annual accounts can be mapped onto one another. If you build software without making that mapping explicit, the bill arrives at the end of the year, when a posting to a cost centre can no longer be traced back to the programme on which the council decided.

Making changes without losing the original position

Once the year is under way, the budget changes. A new task is added, a central government contribution comes in differently, a project slips, a reserve is drawn on. Each of these calls for a decision, and that decision should leave a trail: which proposal, which amounts, which year, who was authorised to approve it and on what date. What goes wrong in practice is that the new position overwrites the old one. After that, nobody can show what was originally approved, and that is precisely the figure an organisation is held to account against afterwards.

The workable solution is a budget built in layers: the position approved at the outset stays untouched, every change is a separate amendment layer with its own attributes, and the current position is the sum of those layers. A report can then show three columns at once: approved at the outset, after amendments, and actual. Delegated authority belongs here too. A shift within a single programme is often an officer-level power; a shift between programmes or an increase to a ceiling is not. Software that knows the difference can process the first kind quickly and block the second until the decision is made, rather than sending everyone through the same heavy procedure.

The commitment is the link that ties all of this to reality. At the moment a contract is awarded, the full amount should be recorded in the system as a commitment, spread over the years in which the work is delivered. On receipt and approval, part of that commitment is released and a cost entry appears. What remains is an outstanding commitment that is deliberately reviewed at year-end: does it carry over to next year, or is it released? That review is one of the few places where an organisation genuinely gets a grip on old, forgotten commitments. If your procurement process still runs separately from your financial administration, the place to start is custom procurement software, because a commitment that only arises at the invoice is a commitment you can no longer steer.

On the invoicing side, a lot will change in the coming years: e-invoicing under the European standard and reporting per transaction rather than per return. What that requires is covered on ViDA software.

What PFM software needs to do

These six components cannot be separated. A budget without commitments gives an overly rosy picture, commitments without a liquidity forecast say nothing about when the money actually leaves, and the whole is unusable if the trail to the source document is missing.

Multi-year budget preparation

Rounds, proposals and scenarios in the same structure as the annual accounts, with all years of the series side by side.

Changes as a separate layer

The opening position stays in place, every change is its own amendment with a decision and date, and the current position is the sum.

Commitments administration

Awarding a contract records the full amount, spread over the years. Receipt releases part of it, and the remainder is reviewed at year-end.

Cash management and liquidity forecasting

Outstanding commitments, awarded contributions and investment planning converted into expected cash flows per period.

Annual accounts and interim reporting

Progress reports and the annual accounts from the same source, with explanatory notes and the periodic provision of information to third parties.

Audit trail per entry

From annual accounts line to entry to source document, recording who approved it and when, without overwriting.

Cash management and liquidity in a public-sector budget

Treasury in a company seeks return within a risk framework. In a public organisation the brief is narrower and stricter: the money must be there when an obligation falls due, at as little risk as possible. For Dutch local and regional government bodies, holding surpluses and investing them at your own discretion is not a free choice, because funds that are not immediately needed are parked with the Treasury. There are also rules on the borrowing side. In practice, an organisation works with a limit on the amount of short-term funds it can raise, the cash-in-hand limit (kasgeldlimiet), and with a standard that caps how much of its long-term debt may be reset in a single year, the interest rate risk standard (renterisiconorm). We do not quote amounts or percentages here, because they differ by type of organisation and by year; what you can expect from software is that these limits are stored as parameters in the system and that you can see where you stand before you take out a loan.

A usable liquidity forecast in the public sector does not come from historical cash flows, but from decisions that have already been made. Outstanding commitments know when they fall due. Awarded grants and contributions have advance-payment rhythms and conditions. Investment projects have a schedule that tells you when instalments are due. On the income side, much is fixed in the periodic rhythm of central government contributions and in the assessment cycle of your own levies. Put those four sources side by side and you get a forecast that can be explained month by month and adjusted when a project slips. If you manage grant flows as the awarding body, this ties in with custom grant management software, where the decision, the conditions and the advance payments belong together.

Cash management also involves a register of loans, and in the public sector that register is rarely purely internal. Alongside the organisation's own loans, there are often on-lent loans to related parties and guarantees issued to third parties. The latter two do not appear as expenditure in the budget, but they do represent a risk that must be explained in the accountability reporting and that can suddenly cost money at the wrong moment. Software that only tracks your own debt therefore misses half the picture. What you want to record, per loan and per guarantee, is the counterparty, the term, the repayment schedule, the security and the expiry date, with an alert before anything falls due.

On the purchasing side, the invoice flow connects to this. An incoming invoice should automatically be matched to an existing commitment, so that the approver only has to confirm that the service was delivered and does not need to work out again which budget it belongs to. Where suppliers invoice electronically, in the Netherlands mostly via the Peppol network or in a structured message format, that matching can be done largely without manual work. This is not a luxury: in a manual flow the delay almost always arises between receipt and the invoice reaching the right approver, and that delay costs a public organisation both statutory payment terms and goodwill with smaller suppliers. More important for this topic, every invoice that is booked outside the commitment punches a hole in the picture your budget holders manage by.

The payment process itself is where most errors are prevented or made. Segregation of duties is not a formality here: whoever creates a supplier or changes a bank account must not also release the payment. A change of account number should have its own approval step, because that is the route by which payment fraud gets in. And because a payment batch in the public sector is rarely signed off by one person, the system must also be able to show who in the chain saw what.

  • Budgeted, committed and actual side by side per budget holder
  • Forecast built from decisions, not from historical averages
  • Cash limit and interest rate risk standard as a parameter, not a loose memo
  • Advance payments and settlement of contributions in the same ledger
  • Segregation of duties between creditor management and payment release
  • Bank account changes with their own approval step
  • Outstanding commitments reviewed deliberately at year-end
Not sure about a large project yet?

Test your idea first: a working prototype in 1 day

With OneDayBuild, we make your idea tangible in a single day for €1,150, so you know whether further development is worth the investment. Decide to go ahead with the full build? Then we deduct the cost in full.

View OneDayBuild →

From budget to annual accounts and compliance

On paper, the chain is simple: an approved budget, a commitment, a delivered service, an invoice, a payment, an entry in the general ledger and finally a line in the annual accounts. In practice, that chain almost always breaks at the same point: the coding. The council decides at programme level, the department works with projects, the finance team books to cost centres and general ledger accounts, and external reporting requires a breakdown by task area. If those four classifications are not linked together in one place, the same manual work reappears every year: a spreadsheet that moves entries to programmes, built by one person, with no trace of the choices made along the way.

Software solves this by recording the coding once and deriving it everywhere else. An entry then gets not just a general ledger account but, in the same record, the programme, the task area and, where relevant, the project. Reporting becomes a matter of aggregating rather than recalculating, and interim reporting comes from the same source as the annual accounts. That is not only more efficient; it is the only way to ensure interim reports and the annual accounts do not contradict each other.

The chain also stands or falls with the monthly close. Accounts payable, accounts receivable, fixed assets, loans and payroll are sub-ledgers that each have their own total, and that total should reconcile to the general ledger. If it does not, the difference moves to a suspense account, and suspense accounts are where problems stay invisible for months. What you should expect from software is that it shows this reconciliation itself, per period, with the open differences listed, rather than someone checking it quarterly in a spreadsheet. This is also the point where an audit stalls fastest: an unexplained suspense account costs more investigation time than all other findings combined.

Public-sector reporting also has a number of topics that business software rarely handles well. Investments are capitalised and depreciated over several years, with a distinction between investments with and without economic benefit that affects how they are covered. Reserves and provisions are not free-floating pots but each have a decision and a purpose, and movements in them should be visible as a separate line rather than as a correction to a programme. Contributions received with a spending obligation are not income until the service has been delivered, so they belong on the balance sheet. And overhead allocation requires an explicit method, because the result per programme shapes the picture on which you are held accountable. These are exactly the areas where a custom-built system adds value or becomes a problem: the logic can be made explainable, but someone has to have designed it deliberately.

Between the budget and the annual accounts lies the year itself, and that is the period in which you can still adjust course. An interim report that only sets actuals against budget delivers its signal too late, because part of the year has by definition not been booked yet. What does work is a forecast per budget: actuals to date, plus outstanding commitments, plus the budget holder's own estimate for the rest of the year. That third element is subjective, which is precisely why the system should record who gave that estimate and when. A variance then becomes a conversation about an expectation rather than a surprise at year-end close, and you immediately have the justification for the budget amendment that will usually be needed.

In public-sector practice, two further things come on top of this and are often forgotten in a design. The first is the separate accountability for earmarked funds: money that comes from another government body for a specific purpose generally has to be accounted for per scheme, with its own definitions and its own reference dates that do not coincide with your programme structure. That means a posting must be linkable to a scheme as well as to a programme and policy field, because filtering afterwards on a description in the journal entry does not work. The second is access. In the social domain and at implementing bodies, payments are tied to individuals, which makes financial data personal data as well. Roles, traceability of who viewed what, and a retention period per type of data then belong in the design, not in the handover.

Lawfulness is the question that comes on top of this and does not exist in a commercial company. Not: did we book it correctly, but: was this expenditure permitted? That breaks down into a few verifiable questions. Did the amount fall within a valid, approved budget? Was the contract awarded in accordance with your own procurement rules and, above the applicable thresholds, the public procurement regime? Were the conditions complied with of a scheme from which payment was made or to a recipient to which funds were awarded? Was the service delivered and confirmed by the right person? Where a board itself issues a statement on financial lawfulness with the annual accounts, that question stops being an afterthought and becomes a requirement on how your processes are set up, and therefore on your software.

The trail an auditor needs to be able to follow

An audit almost always starts at a figure in the annual accounts and works downwards. So the question is not whether your system produces attractive reports, but whether someone can click through from a total, via a selection, to the individual postings, and from a posting to the document and the approval beneath it. If that cannot be done in the system, it happens by email, and then the availability of your staff determines how long the audit takes.

What this requires is limited and concrete. Postings cannot be edited but can be corrected, so the original entry remains alongside the correction. Every approval records who, when and on the basis of what mandate, and that record lives in a log that even an administrator cannot rewrite. Documents are attached to the posting rather than sitting in a network folder, and the commitment, invoice, delivery confirmation and payment are linked to one another. For a sample, the auditor must be able to draw and export a selection themselves, without anyone preselecting the items. Findings, follow-up and their status also need a place, because that is the bridge between two audit years. If that oversight side is your core need, custom audit software is the broader topic; here, the audit trail is a property of the financial administration itself.

Budget authority and mandates Layered budget amendments Commitments per annual tranche Programme and policy area in a single entry Reserves and provisions Accruals and deferrals Cash flow forecast Immutable audit log Sampling and export

When an off-the-shelf package is the better choice

Public finance is no blind spot in the software market. Mature financial packages exist that explicitly target municipalities, provinces, water boards and implementing bodies, with the prescribed programme and policy-area structure built in, commitment accounting as a standard feature, and suppliers who update their product whenever reporting rules change. For some readers of this page, that is the wiser route, and we would rather say so now than halfway through a conversation about custom development.

If your finances follow the usual pattern of budget, amendment, commitment, actuals and annual accounts, you buy that off the shelf, including the maintenance of those rules. Your effort shifts from building to configuring and to keeping your own coding structure clean, which is work you would have had to do anyway.

Custom development pays off in fewer situations than suppliers suggest, but a few come up again and again. The first is the layer around the general ledger: an environment where budget holders, project managers and policy officers see their own position, request amendments and record commitments without having to understand the financial package. That is where an organisation saves the most time, and it is exactly the part that packages do least well. The second is a custom calculation rule: an allocation model, a grant scheme or a funding method that follows from your own policy and does not fit a standard form. The third is connecting to specialist applications, because the figures on levies, benefits, permits or property arise outside the financial administration and must flow in without manual work.

Two practical points often influence the choice more than functionality does. The first is the transition itself: you do not start from zero, but with an opening balance, outstanding commitments, credits that are halfway through their term and a history you still have to account for. How many years you carry over, and in what form, is a decision to make early, because it determines a large part of the work. The second is that a public organisation usually has to tender the contract itself. The requirements you write down then determine the outcome: if you list functions that packages already have, you get packages, and if you describe only the problem, you get proposals that cannot be compared. The choice between a package and custom development therefore belongs before the tender, not after it.

The downside deserves an honest mention. With a custom build, keeping up with changing reporting and accountability rules is your own organisation's responsibility, and that is recurring work, not a one-off cost. That is why the middle road is often the best answer: a package as the administrative core and general ledger, custom development for the portal, the integrations and your own calculation rules, with a documented boundary between the two. You can read how we approach that trade-off under custom ERP systems.

  • Package if your finances follow the usual pattern
  • Package if you want to outsource maintenance for rule changes
  • Custom for the layer where budget holders do their own work
  • Custom for your own allocation, funding or grant model
  • Custom for integrations with your specialist applications
  • Middle road: package as general ledger, custom front end

Frequently asked questions about PFM software

In a business, a budget is an expectation and profit is the yardstick: margin per product or customer, variance against results, consolidation of a group. In a public budget, the budget is a decision with a ceiling per programme, and the question is whether every item of expenditure fell within a valid decision. Two things follow that business software rarely handles well: money is committed at the moment of award, not at invoice, and the accountability afterwards must be traceable for every posting back to a source document and an approval.

Because a long period can pass between awarding a contract and the first invoice. If that award is not recorded anywhere, the budget looks available while it is already spent, and that gap only comes to light when there is nothing left to steer. With a commitments ledger, a budget holder works with three figures side by side: budgeted, committed and realised. When the goods or services are received, part of the commitment is released and the cost posting appears. Whatever is still open at year-end is deliberately reviewed: carry it over to next year or let it lapse.

Yes, and it is a design choice you need to make up front. The workable set-up is a layered budget: the position established at the start stays untouched, each amendment is a separate mutation with a reference to the decision, the amounts per year and the date, and the current position is the sum of those layers. A report can then show the adopted, amended and actual figures at the same time. What you want to avoid is the new position overwriting the old one, because that removes exactly the figure you are held to account against afterwards.

By being drillable from top to bottom: from a total in the annual accounts to the underlying postings, and from a posting to the document and approval behind it. This requires a few properties. Postings can be corrected but not altered, so the original entry always remains. Every approval records who, when and on what mandate, in a log that even an administrator cannot rewrite. Commitment, invoice, delivery confirmation and payment are linked to one another. And the auditor can draw and export a sample themselves.

Often yes, and for some organisations that is the wiser route. Mature financial packages exist for the public sector, with the prescribed structure of programmes and task areas built in, and with suppliers who update their product when reporting rules change. Custom becomes interesting at the layer around it: an environment where budget holders see their own position and record commitments, your own allocation or granting model, or integrations with line-of-business applications where the figures originate. The middle route is common: an off-the-shelf package as the general ledger, custom software at the front.

The source code and documentation belong to the client and are kept in a repository you have access to yourself. Delivery includes a description of the architecture, the data models, the integrations and the deployment, so another party can take it over without first having to reverse-engineer it. We work with mainstream technology, because that determines how many parties can take on the maintenance. For a financial system, this also includes a description of the coding structure and of the calculation rules behind depreciation, cost allocation and coverage.

Related services

Custom procurement software

The commitment arises at the award, so in the procurement process. For request, award, contract and delivery confirmation in a single chain: custom procurement software.

Custom audit software

If your question is mainly about the audit work itself, with work programme, findings and follow-up over the years, then custom audit software is the broader topic.

Custom grant software

If you award funds yourself, the grant decision, conditions, advance payment and final settlement belong together: custom grant software. To reconcile bank transactions and sub-ledgers, financial reconciliation software is the place to start.

A closer look at your budgeting and accountability process

Tell us how your budget is currently put together, at what point a commitment becomes visible in your accounts, and what an auditor could not trace on their own last year. Those three answers usually show whether an off-the-shelf package, custom software or a combination fits best, and where the first gains are.

Edit content