# When compliance reads as failure *Design rules for recording deviations from the oversight plan, together with their causes* **Faizul Abrori** · Auditor, Inspektorat Aceh, Banda Aceh, Indonesia ORCID: 0009-0005-8378-6639 Version 1.0 · 23 September 2026 Translated from the Indonesian original, version 1.3: https://doi.org/10.5281/zenodo.22910151 --- ## Summary An inspectorate's annual oversight work programme is set as an elaboration of the plan for the guidance and oversight of regional government administration, yet some oversight mandates arrive after that programme has been set. An inspectorate that carries out those mandates takes on assignments that are not in its original plan, and by the end of the year its plan-to-actual conformity has fallen. This note argues that the problem lies not in the deviation but in the **loss of the cause** behind it. A variance that keeps no reason can be read in only one way, as planning that missed its mark, even though part of it is in fact compliance with mandates that came later. From there the note sets out five design rules. The aim is not to make deviation harder but to turn it from a figure of failure into information that management can use. The note states the rules and how to measure them, not the results of measuring them. --- ## 1. A figure that makes compliance look like negligence Picture two inspectorates at the end of the year. Both report that seventy per cent of the assignments they carried out were in their original plan. The first dropped thirty per cent of its plan because it ran out of time and staff. The second set aside thirty per cent of its plan because new mandates arrived after the plan was set, and it carried those mandates out. The figures are the same. What lies behind them is opposite. One was negligent, the other compliant. Without further information, no reader of the report can tell the two apart. Because the figure is all there is, the reader will settle on the easiest interpretation, which is that the planning was not careful enough. This is the form the damage takes. Not the deviation itself, but the fact that legitimate and illegitimate deviations are **added up into a single figure**, and that figure is then used to judge the quality of the planning. ## 2. Two kinds of deviation counted as one | | Cause | Could better planning have prevented it? | |---|---|---| | Explained deviation | A mandate, request, or event that arose after the plan was set | **No.** The cause lies outside the planner's control | | Unexplained deviation | The plan was not carried out, and nobody knows why | Yes. This is the one that should be fixed | The first kind genuinely exists, and this can be shown without drawing an example from any particular office. The 2026 plan for the guidance and oversight of regional government administration is set out in the annex to a ministerial regulation issued on 19 January 2026 and promulgated on 2 February 2026, while the annual oversight work programme is an elaboration of that plan (Kementerian Dalam Negeri Republik Indonesia, 2026, Article 1 point 3, Article 5, and Article 6(1)). The plan the programme follows therefore becomes available only after the year it governs has begun. A lag of this kind is built into the order in which the plans are drawn up, not into the way any one office draws up its own. The distinction looks simple until someone tries to make it at the end of the year. By then the assignments are finished, the person who made the decision may have moved on, and the letter that triggered the assignment may be filed somewhere else. Reconstructing reasons a year later produces memories, not records. That is why the distinction has to be made **when the assignment comes in**, not when the report is prepared. ## 3. Why the reasons disappear **The plan is treated as a document, not as a statement of where things stand.** Once a document has been formally adopted, its standing changes. It becomes something to be obeyed rather than something to be managed. Adjusting it then requires a justification, while ignoring it does not. So new assignments are carried out outside the document, and the document is left to become an archive that no longer describes anything. **There is nowhere to write the cause down.** As long as the list of assignments holds only what each activity is, when it takes place, and its budget, there is no place to answer why an assignment is there when it was never planned. Information that has no column depends on individual initiative, and so whether it gets recorded is left to chance. **At the time, everyone knows the cause.** This is the most misleading of the three. When a mandate arrives, everyone working on it knows why it is there, and precisely because of that nobody feels the need to write it down. Shared knowledge feels like a record, but it does not last as long as the documents it refers to. ## 4. What breaks **Compliance cannot be proven.** An inspectorate that set aside its plan to carry out a legitimate mandate has no way to show that this is what happened. All it has is a figure that reads as negligence. **Poor planning is shielded as well.** This is the reverse, and just as bad. Because every deviation looks the same, the deviations that truly stem from weak planning are hidden inside the same figure. **The pattern of causes is never seen.** If mandates from a particular source always arrive in the same month every year, that is information that can be used to leave room when drawing up the next plan. As long as the causes are not recorded, the pattern never shows, and next year's plan is drawn up without knowing it. **Management loses what it needs to decide.** The conformity figure holds no answer to the two questions that shape the next decision: why a plan was set aside, and what needs to change so that it does not happen again. However carefully the figure is calculated, the answers are not in it. ## 5. The design rules **Rule 1. Assignments outside the original plan can be recorded, together with their cause and when they came in.** Not only the fact that they exist, but when they came in and on what basis. The timing matters as much as the cause, because it is what separates mandates that arrived before the plan was set from those that arrived after. **Rule 2. The plan can be adjusted, and each adjustment is recorded.** A plan that cannot be touched risks being abandoned without a trace, while a plan that can be edited freely loses its meaning as a plan. What is needed is a recorded adjustment, so that the original plan can still be read side by side with the plan currently in force. This rule rests on an obligation that already exists, though it does not stop there. Indonesia's government internal audit standards already require the head of each government internal audit unit to review and adjust its plans in response to change, and require those plans, including significant interim changes, to be communicated to the head of the ministry, agency, or regional government for approval (Asosiasi Auditor Intern Pemerintah Indonesia, 2021, paragraphs 2010 and 2020). Approval of significant changes to the oversight plan therefore does not rest with the inspectorate alone. What the standards do not yet set out is the form that approval takes, and that is where this note adds a proposal: that approval should produce a plan that can still be compared with the one before it. **Rule 3. The year-end variance is broken down, not added up.** A plan-to-actual conformity report presents at least two figures, one for explained deviations and one for unexplained ones. A single combined figure answers no question worth asking. **Rule 4. Reasons are written to be read, not to fill a column.** This is the rule that decides whether the other four are useful or turn into an administrative burden. A trail that nobody ever reads constrains nothing, and over time it becomes merely a box to be ticked. A trail that management reads when making a decision becomes an input. What makes the difference is not the form of the entry but whether anyone reads it. **Rule 5. What is measured is not only how much deviates, but when the causes arrive.** Record the interval between the setting of the plan and the arrival of each mandate. Seasonal patterns then become visible, and next year's plan can leave room for the mandates whose timing can in fact be foreseen. ## 6. How to tell whether the rules work 1. What share of the assignments outside the plan have a cause recorded when they came in, rather than reconstructed at the end of the year? 2. Can someone read the plan in force today side by side with the original plan, and see what changed between the two? 3. Does the year-end report present explained deviations separately from unexplained ones? 4. Has any recorded cause ever changed a single decision in drawing up the next plan? If not, Rule 4 is not yet working, however neatly the column is filled. ## 7. Limits of the claims This note states design rules and how to measure them. It does not report measurement results. Both figures in the example in section 1 are invented for the purpose of explanation, not taken from any inspectorate. No regulation requires a history of changes to the oversight plan to be kept, and this note does not claim that one does. Rules 2 to 5 are the author's proposals. They rest on the obligation, set out in the internal audit standards, to adjust plans, but go beyond its wording. This note does not argue that every deviation that comes with a reason is automatically the right decision. Its claim is narrower: a deviation whose cause is recorded can be assessed, while a deviation whose cause is not recorded can only be counted. Assessment remains within management's authority, and this note speaks only to what information should be available when that assessment is made. This note does not discuss any audited entity, does not name any particular mandate or institution, contains no audit results, and does not assess any particular inspectorate. --- ## References Asosiasi Auditor Intern Pemerintah Indonesia. (2021). *Peraturan Nomor PER-01/AAIPI/DPN/2021 tentang Standar Audit Intern Pemerintah Indonesia* [Regulation No. PER-01/AAIPI/DPN/2021 on the Indonesian Government Internal Audit Standards]. https://itjen.pertanian.go.id/storage/master/file/knno0IEe_saipi2021.pdf Kementerian Dalam Negeri Republik Indonesia. (2026). *Peraturan Menteri Dalam Negeri Nomor 4 Tahun 2026 tentang Perencanaan Pembinaan dan Pengawasan Penyelenggaraan Pemerintahan Daerah Tahun 2026* [Minister of Home Affairs Regulation No. 4 of 2026 on Planning the Guidance and Oversight of Regional Government Administration for 2026] (Berita Negara Republik Indonesia Tahun 2026 Nomor 74). --- ## About the author Faizul Abrori is an auditor at Inspektorat Aceh, the internal audit office of the Aceh Provincial Government in Indonesia. The work is mostly reviews and audits of regional government finances. The rest goes into designing and building, personally, the tools used in that work, and into writing about oversight practice from personal experience and views. This note is one of them. Email `halo@fairi.id`. Other writings and a full profile at `https://fairi.id`. > The views expressed in this note are the author's own and do not represent > the official position of Inspektorat Aceh.