How Postwar Public Financial Management Works | Revenue, Payroll, Budget, Treasury, Procurement and Audit

The Deeper Architecture: Public Financial Management Is a Chain of Custody for Public Money

Public financial management is easiest to understand when every unit of public money is treated as something that must move through a visible chain of authority. Money is not legitimate merely because a government possesses it. The state has to show where it came from, which law or appropriation authorised its use, who committed it, who verified the good or service, who approved payment, where the payment went, how the transaction entered the accounts and which independent institution can inspect the record later.

That chain turns cash into public finance. Without it, a government may still spend money, but spending depends increasingly on personalities, emergency discretion and fragmented records. The citizen cannot distinguish a lawful payment from patronage, the treasury cannot distinguish an unpaid commitment from free cash, and the next administration cannot reconstruct what the previous one promised.

A postwar PFM system therefore has a simple deep job: make every important fiscal state legible enough that authority, money, obligation and evidence remain connected from revenue to audit. The sophisticated software can come later. The chain of custody has to exist first.

The Recovery Sequence Begins With Control Before Optimisation

Fragile administrations are often offered advanced reforms too early: performance budgeting, sophisticated forecasting, integrated information systems, accrual accounting, complex procurement analytics or elaborate programme structures. These tools can be useful in capable states. They do not substitute for elementary control.

A recovering government first needs to answer simpler questions reliably. Which bank accounts exist? How much cash is actually available? Which employees are authorised? Which contracts are valid? Which invoices are unpaid? Which taxes may legally be collected? Which office may approve a payment? Which accounting period is still open? Which record is canonical?

The recovery sequence is therefore usually visibility → authority → control → reconciliation → predictability → reporting → optimisation. A state that skips the early stages may produce modern-looking reports on top of financial states nobody trusts.

Start With the Fiscal Authority Map

Before rebuilding systems, map who is legally allowed to do what. Who sets tax policy? Who administers taxes? Who prepares the budget? Who authorises appropriations? Who may move funds between budget lines? Who releases cash? Who signs contracts? Who verifies delivery? Who approves payroll changes? Who operates government bank accounts? Who records transactions? Who audits?

War often leaves overlapping authorities behind. Emergency committees may have acquired spending powers. Regional bodies may collect revenue independently. Ministries may open accounts outside treasury control. Donor projects may operate separate payment systems. Security institutions may use exceptional procedures long after the original emergency has changed.

The authority map does not assume every exceptional arrangement should disappear immediately. It identifies the actual operating state and distinguishes lawful delegation, temporary exception, ambiguity and unauthorised practice. You cannot reconcile money until you can reconcile authority.

The Fiscal Year Is a State Machine

A budget year is not one event. It moves through states: forecast, policy decision, draft budget, legislative authorisation, allotment, commitment, verification, payment, accounting, reporting, audit and correction. Each stage changes what government may legitimately do next.

This matters because a transaction can be economically sensible and still be fiscally invalid if it entered at the wrong state. A ministry may have a budget allocation but no cash release. It may have cash but no procurement authority. It may have a signed contract but no evidence of delivery. It may have received goods but no valid invoice. PFM is the discipline of preventing one valid state from being mistaken for every later state.

A strong system therefore records transitions. Who moved this obligation from proposed to committed? What evidence moved it from committed to payable? What approval moved it from payable to paid? The audit trail is simply the memory of those state changes.

Macro-Fiscal Planning Sets the Boundary Before Ministries Bargain

Every budget begins with an estimate of the resource envelope. Expected tax revenue, customs receipts, non-tax revenue, grants, financing and debt-service obligations create a first approximation of how much room the government has.

Postwar forecasting is unusually uncertain. Tax records may be incomplete, production may have shifted geographically, imports may be volatile, donor commitments may be conditional and large expenditure needs may be only partially known. Precision can become theatrical if the underlying data is weak.

A better approach separates a central estimate from uncertainty. What revenue is highly likely? What is plausible but not secured? Which grants are signed? Which are pledged? Which debt payments are contractual? Which emergency needs could appear? A fiscal plan becomes more robust when it distinguishes money the state can count from money it merely hopes will arrive.

Budget Ceilings Turn Scarcity Into an Operating Constraint

If ministries prepare unconstrained wish lists and only later discover the available envelope, the budget becomes an exercise in cutting promises after expectations have already formed. Expenditure ceilings give ministries an earlier boundary within which to prioritise.

The ceiling should be credible enough that agencies can plan. A ceiling that changes radically every week does not create discipline; it pushes agencies to overbid because nobody believes the first number. Predictable scarcity is easier to govern than arbitrary scarcity.

Ceilings also expose trade-offs. If the health ministry wants more medicines, it may need to postpone equipment. If government wants to expand salaries, it may have less room for maintenance. Budgeting becomes real when every additional promise displaces another use of finite resources.

A Chart of Accounts Is the Grammar of Public Money

Governments need a common classification system for transactions. Which ministry spent the money? For what economic category? On which programme, project, location or funding source? Was it salary, goods and services, transfer, capital expenditure or debt service?

The chart of accounts supplies that grammar. If agencies classify the same transaction differently, national reports stop reconciling. If the classification becomes too complicated for frontline users, error and workaround spreadsheets multiply.

The best postwar chart is therefore not the most intellectually elaborate one. It is the simplest classification capable of supporting legal control, management and reporting while remaining usable across the institutions expected to adopt it. Complexity can be added when the administrative system can carry it.

Commitment Control Stops Tomorrow’s Cash From Being Spent Twice Today

A government creates a financial obligation before it makes a payment. Signing a contract, issuing a purchase order or authorising work can commit future cash. If those commitments are not recorded, the treasury may believe money is still available and authorise a second obligation against the same resources.

This is how arrears can grow inside a budget that appears balanced on paper. Cash accounting shows what has been paid. Commitment control shows part of what government has already promised.

A strong control requires agencies to check budget authority before committing, record the commitment promptly, update it when the amount changes and close it when the obligation is completed or cancelled. The financial state then moves from free balance to committed balance before the invoice arrives.

The Purchase-to-Pay Chain Should Be Explicit

Public procurement becomes safer when the sequence is visible: need identified → budget confirmed → procurement method authorised → supplier selected → contract or purchase order issued → goods or services received → receipt verified → invoice checked → payment approved → cash transferred → transaction recorded → records retained.

Each step exists because a different failure is possible. Without budget confirmation, procurement creates unfunded obligations. Without supplier-selection evidence, favouritism becomes harder to detect. Without receipt verification, government can pay for goods that never arrived. Without payment records, the same invoice can be paid twice.

The chain should be proportionate to risk. Buying emergency drinking water is not the same as building a power plant. But compression of procedure should be deliberate. Removing a step should create a stronger alternative control, not simply a blind spot.

Three-Way Matching Is a Simple but Powerful Control

Before paying a supplier, a finance office can compare three records: what government ordered, what the receiving office confirms was delivered, and what the supplier invoices.

If the purchase order says 100 units, the receipt says 80 and the invoice says 100, the mismatch needs resolution before payment. If quantities match but the price changes, the contract or approved variation should explain why.

This does not eliminate fraud or error, but it forces several independent parts of the transaction to reconcile. In a fragile system, simple controls that are actually performed are more valuable than elaborate controls that exist only in manuals.

Cash Forecasting Converts the Annual Budget Into a Weekly Reality

The annual budget says what government intends to spend over a year. Treasury needs a shorter-horizon view: what cash will arrive this week, this month and this quarter, and what payments will fall due?

Forecasting should distinguish predictable obligations such as payroll and debt service from more variable procurement and project payments. Ministries need to report expected payment timing rather than merely annual allocations. The treasury can then identify weeks when obligations exceed available cash and act before the due date arrives.

Accuracy improves through feedback. Forecast what will happen, compare with what actually happened, classify the error and update the next forecast. Cash management is a learning loop, not a once-a-year calculation.

Payment Prioritisation Needs Rules Before Cash Becomes Scarce

When cash is insufficient, someone decides who waits. Without explicit rules, payment sequencing can become a hidden political market in which suppliers or agencies with better access are paid first.

A recovery government should define priority classes consistent with law and policy: for example, critical salaries, legally unavoidable debt obligations, essential medicines, emergency operations or other commitments whose delay would create disproportionate harm. The exact order is country-specific. The need for a visible order is general.

Exceptions should leave evidence. If a payment jumps the queue, the authority and reason should be recorded. Scarcity is difficult; discretionary scarcity without receipts is corrosive.

Treasury Consolidation Is About Visibility Before It Is About Centralisation

A treasury single account is often described as one bank account, but the deeper objective is unified visibility and control over government cash. Real arrangements can contain a main account plus subsidiary accounts, transaction accounts or legally protected funds.

The design question is whether the finance ministry can see public cash, prevent unauthorised account creation, consolidate idle balances where lawful and reconcile banking records to government books. A nominally central account with unreported side accounts does not solve fragmentation.

Reform should therefore begin with an account census: bank, account number, legal owner, purpose, balance, signatories, funding source and whether the account is still needed. Closing or consolidating accounts comes after the state knows what exists.

Bank Reconciliation Is Where Two Independent Memories Meet

The treasury records what it believes happened. The bank records what actually moved through the account. Reconciliation compares the two.

Differences can be legitimate timing items, fees, returned payments, unrecorded deposits, duplicate transactions or errors. What matters is that unexplained differences do not accumulate indefinitely.

Regular reconciliation is one of the simplest tests of whether the financial system has reality contact. A ledger that never meets the bank statement can become internally consistent and externally false.

Payroll Is a Recurring Contract Between Position, Person and Payment

A reliable salary payment needs three things to remain connected: an authorised position, an eligible person occupying it, and a payment calculated under valid rules. If any leg breaks, payroll becomes unreliable.

Personnel records should identify the establishment position, grade, location, appointment authority and employment status. Payroll records should identify the person, salary components, deductions and payment destination. Changes—appointment, promotion, transfer, leave, suspension, termination—should flow through authorised procedures rather than informal messages.

The monthly payroll run then becomes a reconciliation event. Did the employee list change? Why? Did unusual allowances appear? Did terminated employees disappear? Did bank failures or returned payments occur? The objective is not merely to remove “ghost workers.” It is to make every salary explainable.

A Payroll Cleaning Exercise Needs a Safe Exception Path

Verification exercises can remove invalid records, but they can also wrongly exclude legitimate employees whose documents were destroyed or whose postings changed during conflict. A control without an appeal route can create new injustice.

Records should therefore distinguish verified, pending verification, disputed and invalid states. A pending case should not automatically be treated as fraud. The employee should know what evidence is missing and where the case can be reviewed.

Good payroll reform combines control with due process. The state needs to stop unauthorised payments without making administrative chaos another source of arbitrary deprivation.

Procurement Planning Starts Before the Tender

A procurement system is often judged by the competition stage, but many failures begin earlier. Was the need real? Was the specification unnecessarily restrictive? Was the quantity based on demand? Was maintenance included? Was the delivery site ready? Was there budget and cash?

Procurement planning connects programme need to market reality. A ministry should know what it intends to buy, when, under which method, from what plausible supplier market and with what downstream operating cost.

This matters after conflict because markets themselves may be damaged. A tender with formal competition may still be weak if only one supplier has transport access, if specifications refer to unavailable materials or if payment delays have driven credible firms away from government business.

Emergency Procurement Needs a Return Path to Ordinary Rules

Emergency procurement can legitimately shorten advertising periods, use framework agreements, direct awards or other accelerated procedures allowed by law. The danger is institutional memory: exceptional methods become easier and therefore remain long after the emergency has changed.

Every emergency method should carry an activation condition, authority, documentation requirement, review threshold and sunset or reassessment point. The process should be able to explain why ordinary competition was not feasible for this procurement at this time.

The principle is not “never move fast.” It is “speed must still leave a trail strong enough for later scrutiny.”

Contract Management Begins After Award

Winning a tender does not deliver a bridge, medicine shipment or software system. Government still has to manage milestones, quality, variations, delays, warranties, retention, disputes and final acceptance.

Contract variations deserve special attention because a competitive award can become economically different after repeated changes. The system should record who authorised the variation, why it was necessary, how price and schedule changed and whether the amended contract still represents acceptable value.

The procurement record should therefore survive through completion. Award is the midpoint of the fiscal chain, not the end.

Assets Need a Register After the Invoice Is Paid

When government buys a vehicle, generator, computer, medical device or building, the financial system should not forget the item after payment. The asset now creates custody, maintenance, insurance, replacement and disposal obligations.

An asset register connects the purchase to an identifiable object, location, custodian, condition and useful life. In a postwar environment, even a partial reliable register can reveal equipment that exists on paper but cannot be located, or assets that remain in service but have disappeared from financial records.

Public financial management therefore extends beyond expenditure. It includes remembering what the expenditure created.

Inventory Control Protects Consumables Between Warehouse and Use

Medicines, fuel, food, textbooks and spare parts can be correctly procured and still fail to reach the intended service. Inventory systems track receipt, storage, issue, transfer, loss, expiry and balance.

The financial record and physical record should reconcile periodically. If the accounts say a clinic received supplies, the warehouse and clinic records should show where those supplies moved. Large unexplained differences are a control signal.

This is the point at which PFM meets operations. Money is useful only when the resource purchased reaches the service that justified the expenditure.

Revenue Administration Is a Pipeline, Not a Single Collection Event

A tax system does not begin when cash reaches a treasury account. It begins earlier with law, taxpayer registration, identification of liabilities, filing or assessment, payment, reconciliation, arrears management, enforcement and appeal. Each stage answers a different question: who owes, how much, under which rule, by when, through which channel, and what happens if the record is wrong?

Postwar administrations often have gaps between these stages. A taxpayer may receive an assessment that never appears in the payment system. Customs may collect revenue that reaches a bank account but not the central ledger. Local collectors may use paper receipts that are difficult to reconcile centrally. The repair task is therefore to reconnect the pipeline before adding complexity.

The minimum viable system needs a unique enough taxpayer identity, a lawful liability, a record of payment and a route for dispute. Enforcement without appeal creates arbitrary power. Appeal without records creates endless uncertainty. Revenue administration becomes legitimate when obligation and correction travel through the same visible system.

Customs Revenue Needs Both Border Speed and Fiscal Control

In many recovering states, border taxes and customs duties can be an important early source of revenue because imports are visible at a smaller number of entry points than domestic economic activity. That concentration can make collection administratively easier than rebuilding a sophisticated domestic tax base immediately.

But border collection can also become a high-risk interface. Delays can obstruct essential imports, discretionary valuation can create opportunities for abuse, and fragmented checkpoints can produce multiple charges on the same movement. Customs reform therefore needs clear tariffs, authorised collection points, payment records, reconciliation to treasury accounts and an appeal route for disputed assessments.

The objective is not maximum extraction at the border. It is predictable lawful collection that raises revenue while allowing legitimate trade to move. Fiscal capacity grows when traders can know what they owe before they arrive.

Tax Arrears Need Their Own Ledger

Revenue can be legally due without being collected. If unpaid liabilities are not tracked, the government cannot distinguish a weak tax base from weak collection, nor can it know whether arrears are concentrated among a few large taxpayers or spread broadly.

An arrears ledger should identify the liability, due date, taxpayer, amount, dispute status, collection action and final resolution. Old or legally uncollectable amounts should not remain forever as fictional assets. Genuine arrears should not disappear because collection is politically difficult.

The same principle appears on the expenditure side. A fiscal system becomes reliable when overdue money owed to government and overdue money owed by government are both visible rather than hidden outside the headline budget.

Revenue Forecasting Should Learn From Error

A revenue forecast is useful only if later compared with actual collections. If customs receipts were lower than expected, was trade volume lower, the exchange rate different, exemptions larger, compliance weaker or the original model simply unrealistic?

Forecast error should be decomposed rather than hidden. Repeated optimistic forecasting creates budgets that look balanced at approval and become cash crises during execution. Repeated pessimism can unnecessarily suppress services and investment.

The learning loop is simple: forecast → collect → compare → explain variance → update assumptions. Postwar uncertainty makes the loop more important, not less, because every forecast begins with weaker information.

Tax Expenditures Are Spending Hidden Inside the Revenue System

Governments can support sectors or groups not only by spending money but by reducing tax that would otherwise be due through exemptions, deductions, preferential rates or other provisions. These choices may be justified, but they consume fiscal capacity.

If exemptions are invisible, policymakers can debate every line of expenditure while ignoring large fiscal commitments embedded in the tax code or administrative practice. A basic inventory of major tax preferences helps reveal who receives them, under what authority and whether they still serve the intended purpose.

The principle is symmetry: a government should scrutinise a dollar not collected for a policy purpose with some of the same seriousness as a dollar spent for that purpose.

Supplementary Budgets Are a Correction Mechanism, Not a Second Hidden Budget

Postwar conditions change quickly. A flood, displacement wave, security shock, revenue shortfall or donor decision can make the original annual budget unrealistic. The fiscal system therefore needs lawful ways to revise authorised spending.

Supplementary appropriations, contingency mechanisms and transfers between budget lines can provide that flexibility. Their legitimacy depends on rules: who may propose them, which authority must approve them, what limits apply and how the change enters public reporting.

Flexibility becomes dangerous when the revised budget is effectively written through repeated exceptions. The original appropriation should remain connected to the sequence of authorised changes so citizens and auditors can see how the final spending plan evolved.

Contingency Reserves Buy Time Before a Shock Becomes a Crisis

A contingency reserve sets aside a bounded amount for unforeseen needs. This can be especially valuable during recovery because uncertainty is unusually high and forcing every emergency through a complete annual rebudgeting process may be too slow.

The reserve still needs rules. Eligible uses, approval authority, reporting and replenishment limits should be known. Otherwise a contingency line can become a discretionary fund that bypasses normal prioritisation.

The best contingency mechanism creates speed without invisibility. It accepts that some needs cannot be predicted while refusing to treat unpredictability as permission to abandon the audit trail.

A Medium-Term View Connects Today’s Capital Project to Tomorrow’s Operating Bill

A school, hospital, road, water system or digital platform may require a large one-time investment, but the fiscal commitment continues after construction. Staff, electricity, maintenance, consumables, software licences, spare parts and eventual replacement create recurrent costs.

This is why multi-year expenditure planning matters even when the annual budget remains the legal appropriation instrument. A government should ask not only whether it can build an asset this year, but whether it can operate and maintain the service over the following years.

Ignoring recurrent cost creates a familiar failure: impressive capital projects open and then deteriorate because the operating budget never absorbed them. Fiscal sustainability begins before the ribbon-cutting ceremony.

Public Investment Needs a Gate Before Procurement

Large projects should pass through a decision gate before they become tenders. Is the project aligned with a public priority? Is demand real? Were realistic alternatives considered? Is the design mature enough to cost? Is land available? Are environmental, social and legal requirements addressed? Can the state maintain the asset later?

Postwar urgency can tempt governments and donors to begin construction before these questions are answered. That may create stranded assets, expensive redesigns or facilities that cannot be staffed.

The gate does not need to be bureaucratically elaborate. Its purpose is to stop an attractive idea from becoming an irreversible fiscal commitment before the most important dependencies are visible.

The Vendor Master Is a Small Dataset With Large Consequences

Payment systems need a reliable record of suppliers: legal name, identifier, payment details, status and other information required by law. Duplicate or weak vendor records can create duplicate payments, misdirected transfers and difficulty linking contracts to suppliers.

Changes to sensitive fields such as bank details should be controlled and independently verified. Dormant or invalid vendors should be distinguishable from active ones. The objective is not to create a vast database for its own sake, but to ensure that the supplier named in the contract is the supplier receiving the authorised payment.

Where procurement law requires conflict-of-interest or ownership disclosures, those records should connect to the procurement file rather than live in a disconnected compliance archive.

A Contract Register Makes the Government’s Promises Visible

Contracts create future obligations. A central or federated contract register can identify supplier, contract value, start and end dates, responsible agency, funding source, major milestones, variations and payment status.

This is especially useful when many reconstruction and emergency contracts are active simultaneously. Without a register, finance ministries may see payments only when invoices arrive, too late to understand the full future commitment profile.

The register should not replace the full procurement file. It is the index that makes the portfolio legible enough for cash planning, management and oversight.

Supplier Performance Should Feed the Next Decision

Procurement learns when delivery experience is recorded. Did the supplier deliver on time? Were goods compliant? Were defects corrected? Did variations become excessive? Were warranties honoured?

Performance information should be factual, reviewable and used according to procurement law. It should not become an informal blacklist controlled by personal preference. Suppliers need a route to challenge inaccurate records where those records affect future eligibility.

A government that never learns from contract performance repeatedly pays the cost of forgetting.

Accounting Basis Should Match Administrative Capability

Cash accounting records transactions when cash is received or paid. Accrual accounting recognises economic events such as receivables, payables and asset consumption under a different logic. Each provides different information.

A recovering state should not treat accounting reform as a prestige ladder in which the most complex basis is automatically the best immediate choice. Reliable cash records plus separate commitment, arrears, asset and debt information can be more useful than a nominally advanced system fed by weak source data.

The right sequence is capability first. Build complete transactions, reconciliations and closing discipline; then expand the accounting model as the state can reliably observe the additional economic states the model requires.

Month-End Close Converts Transactions Into a Trusted Period

Financial reporting becomes unreliable when transactions can be posted indefinitely into past periods or when agencies do not reconcile before reports are produced. A month-end close creates a checkpoint.

Typical tasks include bank reconciliation, review of suspense items, payroll reconciliation, recording known commitments or liabilities according to the accounting framework, checking unusual balances and confirming that subledgers agree with the general ledger.

The close does not need to be perfect to be useful. It creates a repeatable moment at which the government asks whether its financial memory still matches reality.

Suspense Accounts Are Temporary Waiting Rooms, Not Permanent Storage

Sometimes a transaction arrives before finance staff know exactly how it should be classified. A suspense or clearing account can hold it temporarily while the issue is investigated.

The danger begins when temporary items accumulate for months or years. Large old suspense balances mean the ledger contains transactions whose real meaning has not been resolved.

Every suspense item should have an age, owner and resolution path. Temporary uncertainty is normal. Permanent unclassified money is a control failure.

Advances Need Liquidation Rules

Government may advance money for travel, field operations, petty cash, emergency purchases or programme activity before final supporting documents exist. That can be operationally necessary.

An advance should remain visible until the recipient provides the required evidence, returns unused cash or otherwise settles the obligation. New advances can be restricted when old ones remain unjustifiably unresolved, subject to fair exceptions for operational needs.

The control principle is that an advance is not an expenditure merely because cash left the treasury. The financial state remains open until the purpose is evidenced.

Expenditure Arrears Need Verification Before Clearance

After conflict, government may inherit a large stock of unpaid invoices and salary claims. Paying everything presented as an old obligation can validate false claims. Refusing everything can bankrupt legitimate suppliers and employees.

An arrears stocktake should therefore classify claims by legal basis, documentation, delivery evidence, age, dispute status and priority. Verified claims can then enter a transparent clearance plan consistent with available cash.

The plan should also repair the mechanism that created new arrears. Clearing the old stock while agencies continue making unfunded commitments simply rebuilds the problem.

Guarantees and Contingent Liabilities Are Promises That May Become Cash Later

A government can assume fiscal risk without borrowing cash directly. It may guarantee a loan, promise support to a state-owned enterprise, enter a public-private arrangement or face legal claims that could require future payment.

These contingent liabilities should be identified and monitored because a shock can convert them into immediate expenditure. A budget that ignores them can appear healthy until several guarantees are called at once.

Visibility does not mean every possible liability should be booked as certain expenditure. It means policymakers understand the exposure, trigger and plausible fiscal consequence before the event occurs.

State-Owned Enterprises Can Carry Fiscal Risk Outside the Core Budget

Public enterprises may deliver electricity, transport, water, fuel, finance or other critical services. Their debts, losses, guarantees and investment plans can affect the state even when their transactions are not part of the central government budget.

A recovering government should therefore know which enterprises create material fiscal exposure, what support they receive, what they owe and which service obligations they perform. Hidden losses can reappear suddenly as recapitalisation needs or unpaid supplier claims.

The ownership question remains distinct from the fiscal-risk question. An enterprise can be operationally separate while still creating obligations the treasury may eventually have to absorb.

Intergovernmental Finance Needs Both Vertical and Horizontal Logic

Transfers to provinces, states or municipalities solve two different problems. Vertical balance asks whether each level of government has enough resources for the responsibilities assigned to it. Horizontal balance asks how resources should differ among jurisdictions with different populations, needs and revenue capacity.

A transparent transfer formula can use variables appropriate to the constitutional and policy context, but the variables need reliable data and should not be changed casually to reward political allies. Transitional grants may be needed while census and revenue information remain weak.

The deeper principle is that decentralised responsibility without predictable finance is not real decentralisation. A local government cannot be accountable for a service if it lacks a credible resource path to deliver it.

Donor-Funded Projects Should Carry Their Future Costs Into the Budget Conversation

An externally funded project can be free to the treasury during construction and expensive afterward. A donor may finance a hospital building while future staffing, utilities, medicines and maintenance become domestic obligations.

Project appraisal should therefore identify counterpart funding and recurrent-cost implications early. The finance ministry needs visibility even when the external funds do not pass through its bank accounts.

Budget integration is not only about recording donor money. It is about ensuring the state can see the full lifecycle of the capability being created.

Monthly and Quarterly Fiscal Reports Are Operating Instruments

Annual accounts arrive too late to manage current execution. Regular reports should show enough information for leaders to see revenue, expenditure, cash, commitments, arrears and material deviations from plan.

The exact format can mature over time. Early reports may be simple but should reconcile to known source systems. A beautiful dashboard built from inconsistent numbers creates faster confusion.

The operating question is always: what decision does this report enable? A cash report should change payment planning. A commitment report should reveal future obligations. An arrears report should identify where the budget is failing to fund its own promises.

An IFMIS Is an Integration Tool, Not a Substitute for Financial Governance

An integrated financial management information system can connect budget, commitment, payment, accounting and reporting workflows. It can improve control by enforcing permissions, preserving audit trails and reducing duplicate data entry.

But software inherits the rules, master data and institutional behaviour placed inside it. If chart-of-accounts design is unstable, user roles are unclear or agencies routinely work outside the system, digitisation can automate fragmentation rather than remove it.

Implementation should therefore treat the system as institutional change: process mapping, configuration, access design, data migration, testing, training, support, reconciliation and controlled retirement of old workflows. The machine is only as authoritative as the governance around it.

Access Control Should Separate Powerful Financial Actions

Digital financial systems make it possible to define who can create, approve, change and pay transactions. Those permissions should reflect organisational responsibility and separation of duties.

One user should not casually be able to create a supplier, enter an invoice, approve it and change the payment account without independent control. Emergency override rights may be necessary, but their use should be logged and reviewed.

Access should also end when jobs change. Dormant accounts and inherited privileges are a common way old organisational states survive inside new systems.

Offline and Manual Fallbacks Are Part of Digital Resilience

A postwar financial system may operate with unstable electricity, connectivity or banking links. If every payment process stops when one network fails, digitisation has created a new single point of failure.

Fallback procedures should define what may continue manually, who may authorise it, how transaction numbers are controlled and how records are entered and reconciled when systems return. The fallback should be bounded enough that it does not become the permanent route around digital controls.

Resilience means the state can continue critical financial operations during an outage and return to one canonical ledger afterward.

Business Continuity Protects the State’s Ability to Pay

Finance ministries and treasuries are critical infrastructure. Fire, cyberattack, disaster, civil unrest or loss of a data centre can interrupt payroll and essential payments.

Continuity planning should identify critical functions, alternate locations or systems, backup data, emergency signatories, banking contacts, priority payments and recovery procedures. Backups matter only if they can actually be restored and reconciled.

The final test is operational: can the government still pay the most critical lawful obligations and preserve a trustworthy record when the normal route fails?

Records Retention Completes the Audit Trail

Financial accountability often happens months or years after a transaction. Contracts, approvals, invoices, payroll changes, bank records, reconciliations and audit responses therefore need retention rules.

Records should remain retrievable for the period required by law and oversight needs, with stronger protection for documents that contain sensitive personal or security information. Digital records need migration plans so a file format or system retirement does not erase institutional memory.

A public-finance system is complete only when the transaction can still be explained after the people who processed it have moved on.

Failure Modes: Public Finance Usually Breaks at the Interfaces

PFM failures are often described as isolated defects: weak budgeting, poor procurement, late reporting or payroll fraud. In practice, the most damaging failures frequently sit between systems. The budget authorises something the treasury cannot fund. Procurement signs a contract the commitment ledger never records. Human resources removes an employee while payroll continues paying. The bank executes a transfer the accounting system never posts. Audit identifies a weakness that management never closes.

This is why repair should follow the transaction across institutional boundaries rather than inspect each office in isolation. A ministry can have a good internal procedure and still participate in a broken end-to-end chain. The question is not merely whether each component exists, but whether the output of one component becomes a trusted input to the next.

The following failure modes are useful because each points to a different first weak link. Fixing the visible symptom without repairing that link usually makes the same problem return in another form.

Failure Mode 1: The Budget Is Aspirational Rather Than Executable

A government may approve expenditure far above realistic revenue and financing. The budget then stops functioning as a constraint. Ministries learn that appropriations do not predict what they can actually spend, so they lobby for cash releases rather than plan against the budget.

The repair begins upstream with a credible resource envelope, realistic ceilings and transparent treatment of uncertain financing. During execution, cash plans should show whether the approved budget remains deliverable or needs lawful revision.

An executable budget does not promise that every allocation will be spent exactly as planned. It creates a credible relationship between legal authority, expected resources and operational delivery.

Failure Mode 2: Cash Rationing Becomes an Invisible Second Budget

When cash is scarce, the treasury may decide which ministries and suppliers receive payment. If these decisions are opaque, the cash-release process effectively rewrites the legislature’s budget without the same visibility.

The repair is to publish or internally formalise prioritisation rules, link releases to cash forecasts and record material exceptions. Where the annual budget cannot be financed, the government should revise it through the lawful mechanism rather than maintain a public budget and a hidden cash budget indefinitely.

Cash management should implement fiscal policy, not quietly replace it.

Failure Mode 3: Commitments Are Invisible Until the Invoice Arrives

If ministries can sign contracts without recording commitments, the treasury sees obligations too late. The same budget balance can appear available to several purchasers, and suppliers discover the shortage only after delivering goods or services.

The repair is commitment control at the point where government becomes legally or administratively obligated. Procurement, budget and treasury systems need one shared view of the obligation before payment becomes due.

The important metric is not only how much cash has been spent. It is how much of the remaining budget is already spoken for.

Failure Mode 4: Payroll Cleaning Removes Real Workers Along With Bad Records

A rapid verification exercise can reduce duplicate or unauthorised payments but also exclude legitimate workers whose documents were lost, whose postings changed during war or whose personnel records were never updated.

The repair combines verification with status categories and appeal. Suspicious is not the same as proven invalid. Cases should be resolvable through documented review, and corrections should flow back into both HR and payroll records.

A control that saves money by making lawful salaries arbitrarily inaccessible can weaken the state it was meant to strengthen.

Failure Mode 5: Emergency Procurement Becomes the Normal Market

Emergency methods are attractive because they are faster. Once organisations become accustomed to them, every urgent request can begin to look exceptional enough to avoid normal competition or review.

The repair is not to ban emergency procurement. It is to define the trigger, authority, documentation, value thresholds, review and sunset. As markets and institutions recover, the proportion of procurement using extraordinary routes should fall unless new emergencies justify otherwise.

Temporary discretion needs an exit condition just as temporary political institutions do.

Failure Mode 6: Audit Produces Reports but No Correction

An audit system can publish excellent findings and still have little effect if management responses are not assigned, deadlines are not tracked and repeated weaknesses carry no consequence.

The repair is a follow-up loop: finding → responsible manager → agreed response → due date → evidence → verification → closure or escalation. Some findings require policy or legislative action rather than administrative correction, and the routing should reflect that.

Audit becomes institutional memory only when the next transaction is less likely to repeat the same defect.

Failure Mode 7: Donor Money Exists Outside the Government’s Fiscal Picture

A ministry may plan a new clinic while a donor is already financing one nearby. A government may assume it needs to fund equipment that an external project has purchased. Or a donor-built system may create future operating costs that never entered the budget forecast.

The repair is information integration even when financial integration is not yet possible or appropriate. Projects, funding commitments, locations, timelines and recurrent-cost consequences should be visible enough for national planning.

The aid-management system owns coordination; PFM owns the question of how external resources alter the government’s fiscal plan and future obligations.

Failure Mode 8: Local Transfers Become Political Favour

When municipalities do not know how transfers are calculated or when they will arrive, local planning becomes dependent on relationships with the centre. Similar jurisdictions can receive very different treatment without a visible reason.

The repair is a transparent allocation rule, predictable payment calendar where possible, reliable denominator data and a correction mechanism for errors. Transitional discretion may be necessary when data is damaged, but the criteria should still be explicit.

Fiscal decentralisation becomes credible when local governments can predict enough of their resource envelope to make and defend real choices.

Failure Mode 9: Software Is Installed Before the Rules Stabilise

A new financial platform can create the appearance of reform while users continue operating through paper, spreadsheets and unofficial workarounds because the configured process does not match legal or operational reality.

The repair begins with process and authority mapping. Decide the budget states, chart of accounts, approval rights, supplier rules, payroll interfaces and reporting requirements before automating them. Pilot difficult interfaces and reconcile new-system outputs to trusted external records.

Technology should compress a working control system, not hard-code an unresolved argument.

Failure Mode 10: Several Numbers Are All Called Official

One office reports one payroll total, another reports a second and a donor dashboard reports a third. Each may be internally reasonable because it uses a different date, population, accounting state or exchange rate.

The repair is canonical ownership plus metadata. Which dataset answers which question? What is the cut-off date? Does the figure represent approved budget, commitment, payment or cash? Are arrears included? What exchange rate applies?

Reconciliation does not require every number to be identical. It requires every legitimate difference to be explainable.

Failure Mode 11: Arrears Become an Unauthorised Financing System

When government routinely buys now and pays suppliers months later, arrears become a form of forced lending. Suppliers may raise prices to compensate, stop bidding or depend on political connections to get paid.

The repair requires both stock and flow. Verify and clear old arrears according to a transparent plan, while strengthening commitment control and cash planning so new arrears do not accumulate faster than old ones are paid.

A clearance programme without prevention is merely refinancing dysfunction.

Failure Mode 12: Capital Projects Arrive Without Maintenance Money

Reconstruction naturally rewards visible new assets. Maintenance is less visible and often easier to postpone. The result can be new roads, generators, clinics or digital systems whose performance declines rapidly after handover.

The repair is lifecycle costing before approval and explicit operating-budget ownership before commissioning. Maintenance plans need responsible institutions, realistic supply chains and recurring finance.

A public investment is not completed when construction ends. It is completed when a sustainable service can begin.

A Postwar PFM Recovery Ladder

Not every state should attempt the same reform package at the same moment. A useful recovery ladder separates capabilities that must exist early from capabilities that become valuable only after the foundations are reliable.

Stage 0: Contain Immediate Fiscal Failure

Protect the minimum functions needed for the state to operate: critical payroll, essential service payments, basic revenue collection, secure government bank accounts and emergency procurement with receipts. Freeze or review clearly unauthorised commitments where lawful.

The objective is continuity without pretending normal control already exists. Every exceptional procedure should be documented because Stage 0 is temporary by design.

Stage 1: Establish Visibility

Inventory cash accounts, payroll, active contracts, arrears, major revenue streams, debt service, donor-financed projects and large fiscal risks. Identify the canonical owner of each dataset.

The first output is not a sophisticated reform strategy. It is a credible picture of the state’s financial position and uncertainty.

Stage 2: Restore Minimum Viable Control

Clarify fiscal authority, budget ceilings, payment approval, payroll change control, procurement methods, bank reconciliation, basic accounting and records retention. Begin commitment control for the most material spending.

The test is whether money can move through a repeatable authorised chain rather than through exceptional personal intervention.

Stage 3: Build Reliability

Improve cash forecasting, monthly close, arrears management, vendor and personnel masters, contract management, regular fiscal reporting and audit follow-up. Reduce the number of unexplained differences among systems.

Reliability means users begin to trust the system enough that bypassing it becomes less attractive than using it.

Stage 4: Integrate

Connect budget, treasury, procurement, payroll, debt, donor information and local transfers through shared identifiers, interfaces and reconciliation rules. Digital systems can now reduce duplication because the underlying states and owners are better defined.

Integration should preserve accountability boundaries. Sharing data does not mean every institution receives authority to change every record.

Stage 5: Improve Performance

Once basic control and reliability exist, the government can use richer analysis: multi-year expenditure planning, public-investment appraisal, programme performance, fiscal-risk analysis, procurement analytics and stronger public reporting.

The state can now ask not only “Was the money spent lawfully?” but “Did the spending produce the public capability it was meant to create?”

Stage 6: Normalise and Retire Emergency Architecture

Close temporary accounts, sunset exceptional procurement routes, absorb donor-supported financial functions into durable institutions where appropriate, archive transitional records and return spending to ordinary legal processes.

The system succeeds when extraordinary controls are no longer required for routine government to remain trustworthy.

Stress Test 1: Revenue Falls Suddenly

Suppose customs receipts or domestic taxes fall well below forecast. Which payments are protected? How quickly does the treasury know? Which expenditure can be delayed lawfully? When does the annual budget require revision?

A resilient system has a cash prioritisation rule, realistic contingency planning and a political route for changing spending rather than allowing unpaid bills to become the default adjustment mechanism.

Stress Test 2: A Large Donor Disbursement Is Delayed

Projects may already have procurement commitments when expected external financing does not arrive on schedule. The system should know which obligations are legally government liabilities, which can pause, and which recurrent costs have already entered domestic budgets.

The test exposes whether pledged, committed and received financing were being treated as the same financial state.

Stress Test 3: The Main Banking Channel Fails on Payroll Day

Can critical salaries be processed through an authorised fallback? Are payment files and beneficiary records protected? How will rejected and duplicate payments be prevented when the normal channel returns?

The objective is continuity plus reconciliation. Emergency payment must not create a second payroll whose transactions never return to the canonical record.

Stress Test 4: Prices or the Exchange Rate Move Sharply

A large price shock can make procurement budgets inadequate and alter the domestic value of foreign-currency debt, grants or contracts. The system needs a rule for reforecasting rather than pretending nominal appropriations still purchase the same capability.

Contract variations, contingency use and supplementary budgeting should remain authorised and traceable even when prices move faster than the annual budget cycle.

Stress Test 5: Displacement Suddenly Changes Service Demand

A municipality may receive thousands of new residents while the transfer formula still reflects old population data. Schools, clinics and water systems experience the change before the national statistical system can fully update.

A resilient transfer system needs a temporary adjustment mechanism with evidence and later reconciliation. Otherwise local fiscal capacity remains attached to yesterday’s population while today’s service demand moves elsewhere.

Stress Test 6: A Major Payroll Irregularity Is Discovered

The discovery should trigger containment without collapsing the whole payroll. Which records are affected? Can questionable payments be isolated? How are legitimate employees protected? Which control failed upstream?

The recovery path is investigate → classify → correct → reconcile → review control → verify next payroll. Public accusations should not substitute for evidence, and evidence should not disappear because politically important people are involved.

Stress Test 7: A New Emergency Requires Immediate Procurement

Can the state buy quickly without losing the transaction trail? Are emergency methods already authorised? Can supplier selection, receipt verification and payment approval be compressed without becoming the same person or office?

The test is not whether normal procedure can remain unchanged. It is whether accelerated procedure still preserves enough independent evidence for later review.

Stress Test 8: Political Leadership Changes

A new government may change priorities while inheriting contracts, payroll obligations, debt service and legally authorised budgets. The financial system should distinguish discretionary future policy from obligations that already exist.

Canonical records and legal provenance protect the transition from becoming fiscal amnesia. New leaders can change policy through lawful mechanisms without pretending old commitments were never made.

Stress Test 9: A Court Invalidates a Tax or Spending Measure

The finance system needs a correction path. Which collections stop? Are refunds or adjustments required? Which expenditure authority is affected? Does the budget need amendment?

Legal review becomes part of fiscal resilience when the system can absorb an adverse decision without treating it as an administrative impossibility.

Stress Test 10: The Financial System Goes Offline

Can critical operations continue under a bounded fallback? Is the last trusted data state known? Are backups recoverable? Can manually authorised transactions be entered later without duplication?

The test should be practised before an outage. A disaster-recovery plan that has never restored the ledger is still a hypothesis.

The Operator’s PFM Checklist

• Can the government state its current cash position with a known cut-off date?
• Can it identify all material government bank accounts and lawful signatories?
• Does every major revenue stream reconcile from assessment or collection to treasury receipt?
• Are revenue arrears and expenditure arrears separately visible?
• Is the annual budget based on a realistic resource envelope?
• Are supplementary budgets, transfers and contingencies lawfully controlled?
• Are commitments recorded before invoices become due?
• Can the treasury forecast major cash needs over the next weeks and months?
• Are payment-priority rules visible when cash is insufficient?
• Does payroll reconcile position, person and payment?
• Can legitimate workers challenge an incorrect payroll decision?
• Does procurement connect need, budget, supplier selection, receipt and payment?
• Are emergency procurement routes temporary and reviewable?
• Can government see its active contracts and major variations?
• Do asset and inventory records remember what expenditure created?
• Are bank and ledger balances reconciled regularly?
• Are suspense accounts, advances and old unreconciled items actively cleared?
• Are debt, guarantees and major state-owned-enterprise risks visible?
• Do local governments receive resources through an understandable rule?
• Are donor-funded projects visible with their future recurrent costs?
• Do fiscal reports enable decisions rather than merely satisfy reporting calendars?
• Does the financial system enforce roles without creating unusable bottlenecks?
• Can critical payments continue during a system or banking outage?
• Do audit findings have owners, deadlines and evidence of closure?
• Can an important transaction still be reconstructed years later?

The First Weak Link Test

When public money fails to reach its intended receiver, begin at the outcome and walk backward. Was the supplier or worker paid? If not, was a payment approved? If not, was the obligation verified? If not, was delivery recorded? If not, was there a valid contract or employment state? If not, was authority and budget available?

The same logic works in the other direction. Start with revenue and follow it forward: lawful liability → collection → deposit → reconciliation → budget authority → commitment → delivery → payment → accounting → reporting → audit → correction.

The earliest broken interface often explains several later symptoms at once. If commitments are not recorded, cash forecasts fail, arrears grow, suppliers lose trust and fiscal reports understate obligations. Repairing commitment control can therefore improve several downstream states simultaneously.

The Complete Mechanism Chain

lawful revenue authority → taxpayer or payer record → assessment or collection → treasury receipt → bank reconciliation → resource forecast → budget ceiling → appropriation → allotment → commitment → procurement or payroll event → verification → payment approval → cash transfer → accounting → period close → fiscal report → audit → management response → correction → institutional learning.

The chain does not imply that every country must use identical institutions or software. It describes the information and authority that must remain connected. Different constitutional systems can distribute those jobs differently while preserving the same logic of custody, evidence and correction.

That is the threshold between a government that spends and a government that can account for spending as an institution.

Public financial management works when every public promise can travel from authority to money to service—and leave enough evidence to travel back again.