Una serie de Pavitt Public Finance
Rules, sin filtros.
Hacer bien la contabilidad, demostrar que el gasto siguió las normas, y las personas que realizan el trabajo.
Un conjunto de ensayos y publicaciones breves sobre los aspectos menos vistosos del cumplimiento normativo y la contabilidad en el sector público. La columna principal, The Compliance Check, se publica semanalmente; IPSAS, Unfiltered continúa dentro de ella como columna. Publicado en Substack, replicado en LinkedIn y archivado aquí para que los patrones sean fáciles de encontrar.
Entradas archivadas
Publicaciones recientes desde el campo.
Un patrón de trabajo sobre lo que ha ido surgiendo en distintas implementaciones. Cada entrada enlaza a la publicación en Substack.
- RU-0162026.10
The binder on the shelf.
Ask a ministry for its competency framework and the usual answer is a search that turns up nothing, or a job description thirty years old. Where a framework does exist, a donor project usually paid for it, a workshop validated it, and it went into a binder in the HR office, because nobody in that chain is paid for it being used the next year. A framework gets used when it describes work a supervisor can observe, rates a handful of skills properly rather than forty by guesswork, plugs into the review and the training plan that already happen, and can be changed by someone in the office.
competencies · appraisal · BenchLeer en Substack → - RU-0152026.10
Did the training move anyone?
The summary table for a training component counts courses, days and participants, and cannot say whether anyone can now do something they could not do before. The count is reported because it cannot fail; an end-of-course test measures what people remember on the last afternoon. Answering the real question takes three things: written levels for what each position needs, a baseline rating by the supervisor from actual work, and the same rating six months to a year later. Measure against the position as well as the person, because a trained officer who is transferred takes the capability with them.
training · capability · BenchLeer en Substack → - RU-0142026.09
The gap analysis you already paid for.
Almost every accrual transition starts with a gap analysis, organised standard by standard, which sizes the task and gets the programme funded, and then usually nothing follows. A list of absences has no order and no names, and nobody's job description has an accounting standard in it. The more useful gap analysis arrives at the end of the first year and is called the audit report: tested against the records, filtered to what is material, and due again next year. Rebuild the remaining plan around the findings, organise it by who does the work, and take first whatever unblocks everything else.
accrual · gap analysis · auditLeer en Substack → - RU-0132026.09
The auditors were trained last.
An audit office is often examining a government's first accrual statements while it is still learning how, and the report is read as though that were not so. The cause is structural: the transition is a finance ministry programme, with that ministry's budget line and steering committee, and the audit office is a separate institution nobody sits at the centre of. Where it stalls is the budget, because audit training is not a line in the programme. The first opinion after a transition measures two things at once, whether the accounts were ready and whether the people examining them were.
audit · accrual · capacityLeer en Substack → - RU-0122026.09
What the whole of government actually owes.
The consolidated statement is the last one prepared and the first one read, because it answers what the state has and what it owes. Adding the entities up is the easy part; the subtractions are where it becomes difficult, since a transaction between two entities was recorded twice, by different people, in different months and sometimes at different amounts. The first attempt is less an accounting exercise than a discovery exercise: undisclosed SOE guarantees, unpaid power bills nobody was tracking. Start with the list of entities and a written reason for each being in or out, then publish it.
consolidation · controlled entitiesLeer en Substack → - RU-0112026.09
Who writes the management response.
Every audit report contains two documents, and the one people work on for the next year is the second. The management response is printed alongside the findings, so an oversight committee, a donor or a new minister reads the finding and the answer together. Answer each finding with a step, a date and a named post and the report reads as a second-year work programme; answer "noted" and it reads as a filing exercise. Disagreeing is allowed, provided the refutation is disciplined. Who writes it is decided months before the report arrives, by whether anyone asked.
audit report · accrualLeer en Substack → - RU-0102026.09
The first opinion is going to be qualified.
The first audit opinion after an accrual transition is almost always qualified, for reasons that were visible three years earlier. The opening balance has never been audited by anybody, because until that year it did not exist in that form, fixed assets are short and the employee benefit liability has never been valued. Saying so at the start of the transition costs nothing and turns the audit report from a blow into a document everyone expected, and a qualified opinion is a working list of specific matters, which is a different thing from a disclaimer.
accrual · audit opinionLeer en Substack → - RU-0092026.08
What the register says is there.
An asset register is wrong in both directions at once: things on the list that left the building years ago, and things in the building that were never written down. Purchases reach the list on their own, because buying creates paperwork somebody is required to file. Disposals need a committee, an inspection and a signature, and transfers need nothing at all. Nobody is measured on whether the register is right, which is survivable until accrual turns it into one of the largest lines on the statement of financial position.
fixed assets · accrualLeer en Substack → - RU-0082026.08
What the government thinks it has.
A treasury can pay every bill on time and still hold only an estimate of what it has in the bank. The ledger balance and the bank's differ for reasons that are dull individually and compound with age, and the only process that settles them runs monthly at best, from the most junior desk in the office. Falling years behind makes no noise at all, and cash rationing means the estimate is what this week's payment decisions actually run on.
cash · reconciliationLeer en Substack → - RU-0072026.08
Three lists, one payday.
The largest recurring expenditure in the budget reaches the ledger as the least informative entry of the month: gross, deductions, net. Everything that made those totals happened in offices that never post a journal, the establishment register, the appointment letter, the acting allowance nobody stopped. An appointment is a commitment that runs for years and clears no commitment control.
payroll · commitmentLeer en Substack → - RU-0062026.08
The contract the ledger never saw.
A contract signed on a Wednesday: three years, four milestones, five per cent retention, a performance security lodged before any work starts. The finance system hears none of it until the first invoice lands eleven weeks later, one number. The largest obligations are the ones the ledger learns about last.
procurement · commitmentLeer en Substack → - RU-0052026.07
The check in the wrong place.
The purchase order clears, because there is money left in the vote. Eight weeks later the invoice arrives and the contract was signed by someone who did not hold the authority to sign it. Budget availability is the part of control a system can always answer. The procedure is the part it usually cannot.
commitment · procedureLeer en Substack → - RU-0042026.07
One chart, three readers.
One invoice for a generator, read by three people who never meet: the budget office reading control, the accountant reading IPSAS, the statistician reading GFS. Three true descriptions of one machine, which only become a contradiction when the chart forces them through a single field.
IPSAS · GFS · budgetLeer en Substack → - RU-0032026.07
Where your law and the standard disagree, the law wins.
The standard says capitalise the truck. The financial instructions, written twenty years before anyone had heard of accrual, say it is expenditure of the vote. Both sentences are on the same desk. What the officer who signs can actually defend, and what has to be written down on the day.
IPSAS · statuteLeer en Substack → - RU-0022026.07
The paragraph that doesn't exist.
Ask a general-purpose AI tool a public-sector accounting question and it will often cite a standard and paragraph that do not exist. Why, in this work, the citation is the part the auditor checks, and the discipline of re-reading every reference against its source before it is marked as supported.
IPSAS · citationsLeer en Substack → - RU-0012026.07
Every rule that binds it.
One ordinary payment answers to three rulebooks at once: the accounting standard, your own law, and the donor's manual. The first post sets out what Rules, Unfiltered is for, and the checking discipline behind every finding.
IPSAS · statute · donorLeer en Substack →
Las publicaciones se editan en Substack para que los profesionales puedan suscribirse y leerlas en su propio feed. Esta página archiva el índice para que los patrones sean fáciles de encontrar.
Acerca de la serie
Patrones, no opiniones.
Rules, Unfiltered está dirigido a profesionales que siguen el cumplimiento normativo y la contabilidad del sector público desde dentro y fuera del detalle cotidiano: responsables de cumplimiento, subcontadores generales, socios de firmas de auditoría, asesores de PFM del lado donante, economistas de ministerios de finanzas. Los artículos abordan patrones recurrentes en distintas implementaciones, no casos específicos de proyectos.
Publicado también en LinkedIn para el feed y los comentarios. Archivado aquí para que los patrones sean fáciles de encontrar cuando alguien busca un precedente en lugar de desplazarse.
Publicado primero en inglés, con traducciones por lotes al español, francés, portugués, ruso y árabe de forma continua. Las traducciones son revisadas, no generadas automáticamente.
Suscríbase.
Principal en Substack, también en LinkedInSuscríbase y cada nuevo artículo llegará a su bandeja de entrada: uno por semana, en lenguaje sencillo, sobre cómo registrar correctamente la contabilidad, demostrar que cumple con sus propias normas y las personas que realizan el trabajo. El archivo permanece aquí; los textos llegan a usted.
Van juntos
Drills + Desk + Series.
La serie aborda los patrones a nivel de sistema. Drills desarrolla reflejos diarios ante los modos de fallo que esos patrones ponen de manifiesto. Desk se ocupa de los casos difíciles cuando aparecen en su contabilidad. Cada uno funciona por separado; juntos cubren todo el ámbito.