A shop sells something. The till records it. At the end of the quarter the owner opens a spreadsheet, works out what is owed, logs into a government portal, and types the figures in.
Every step in that sequence except the last two was already digital. The information existed. It was simply in the wrong place, so somebody had to move it by hand.
That gap is what the OECD’s 2020 report, Tax Administration 3.0, sets out to close.
This article explains what the stages of tax administration actually look like, why digitising existing processes is not the same as transforming them, and what the distinction means for an administration deciding where to invest.
From paper to portal
Tax administration began on paper. Taxpayers filled in forms, delivered them, and clerks processed them by hand. Records sat in cabinets. Calculations were done with a pen.
Digitisation changed the materials. Forms became fields on a screen. Cabinets became databases. Payment moved to a bank transfer. Most administrations have spent the last two decades on this work and much of it has been genuinely useful.
What it did not change is who assembles the information. In both versions the taxpayer gathers their figures, puts them into the format the administration requires, and submits them. The paper version asked for handwriting. The digital version asks for typing.
Tax Administration 3.0 proposes something different. Instead of the taxpayer bringing information to the administration, the administration connects to the systems where the information is already recorded. The OECD describes the result as taxation becoming a more seamless and frictionless process over time.[1]
In the shop example, the till reports the sale. Nothing is exported, re-keyed, or reconciled at quarter end, because the transaction was already captured at the point it happened.
Why the OECD separates digitising from transforming
Administrations have made significant progress over the past decade in taking long-standing practices and making them digital. The OECD treats genuine digital transformation as a separate thing: rethinking how tax is calculated and paid so that those steps happen inside the software taxpayers already run.[2]
An e-filing portal sits on the first side of that line. It is a real improvement on paper and it is still a system the taxpayer has to leave their own records to visit.
The OECD attaches a warning. On the current path, taxpayer burden may rise rather than fall, because taxpayers end up lifting digital records out of systems that already hold them and reshaping them for the administration.[2]
The shop owner keeping paper books and filing a paper return ran one process. Keeping digital books and re-keying into a portal, they now run two.
What 3.0 requires
The 2020 report sets out building blocks that have to exist before seamless taxation is possible. Later OECD work has looked at several of them in detail.
Digital identity. Secure digital identification is a prerequisite rather than a feature. It is what allows a trusted remote connection between a taxpayer, their representatives, and the administration, and it becomes considerably harder across borders.[3]
Connection with natural systems. The systems of administrations, taxpayers and businesses have to be connected so that data moves automatically through machine-to-machine processes, including in real time where that is relevant.[4]
Earlier approaches to data sharing worked differently. Large periodic flows moved between parties and were analysed in pools afterwards. The 3.0 model puts more weight on how timely and how relevant the data is to the taxation process itself.[4]
Taxpayer touchpoints. Friction cannot be removed entirely. Where it remains, the points at which taxpayers deal with the administration still matter and have to be designed rather than inherited.[5]
Electronic invoicing. Where transaction data is produced in a standard machine-readable form at the moment of sale, much of what an administration currently asks for arrives without being requested. The OECD has examined the state of play on this as part of the same programme.[6]
Whether a stage can be skipped
For an administration starting from a low base the question is whether it can go straight to the third stage. The OECD does not address this directly, so what follows is a reading rather than a finding.
Some building blocks do not appear to depend on an extensive estate of earlier digital systems. Digital identity and electronic invoicing are the clearest cases. An administration that has not spent fifteen years building portals also has no fifteen-year-old portal architecture to unwind.
The register is a different matter. Connecting to a taxpayer’s own system requires knowing who that taxpayer is, what they are registered for, and what they owe. Where the register holds businesses that closed years ago, automatic connection produces confident answers that happen to be wrong.
The TADAT framework runs on the same dependency: the integrity of the registered taxpayer base sets a ceiling on what every downstream area can achieve. Automation does not lift that ceiling. It raises the cost of ignoring it.
So the sequencing question is less about which stage comes first, and more about data integrity coming before automation of any kind.
What this means for tax administrations
4 conclusions follow.
Be specific about what a project is. Online portals, real-time integration and a consolidated view of the taxpayer are all worth building. Most of them belong to the digitisation stage. Describing them as transformation sets an expectation of frictionlessness they will not meet alone.
Measure burden, not only capability. A digital process that requires taxpayers to re-enter information they already hold has relocated the cost rather than removed it. The OECD treats that as a live risk on the current path.
Several building blocks sit outside the tax administration. Digital identity in particular is rarely something a revenue authority builds alone. It runs across government, and how quickly one administration can move usually depends on decisions taken elsewhere.
Automation is less forgiving of a weak register than manual processing is. A manual process has a person in it who may notice that something looks wrong. An automated one catches only what it was built to catch.
The move from paper to digital is well advanced in some administrations and still under way in others. What the OECD describes next is not more of it. It runs the other way, and the practical difference is whether an administration is planning for that or assuming it has already started.
Common questions about tax administration digital transformation
Tax Administration 3.0 is a vision for the future of tax administration published by the OECD Forum on Tax Administration in December 2020. It describes a model in which taxation processes are built into the natural systems that taxpayers already use to run their lives and businesses, so that tax becomes a more seamless and frictionless process rather than a separate compliance exercise.
Digitisation moves an existing process onto digital infrastructure, such as replacing a paper return with an online form. Digital transformation rethinks the process so that calculating and paying tax happens inside the software taxpayers already use. The OECD warns that digitisation without transformation can increase taxpayer burden, because taxpayers have to move records out of systems that already hold them to fit the administration’s.
The 2020 report identifies several core building blocks for seamless taxation. These include effective digital identity, the automated connection of tax administration systems with taxpayers’ natural systems through machine-to-machine data flows, well-designed taxpayer touchpoints where friction remains, and electronic invoicing.
Natural systems are the software and devices taxpayers already use in the ordinary course of their activities, such as accounting packages, point-of-sale systems, and banking platforms. Tax Administration 3.0 involves connecting to those systems directly so that taxation happens within them, rather than requiring taxpayers to transfer information into separate government systems.
Some building blocks, such as digital identity and electronic invoicing, do not appear to require an extensive estate of earlier digital systems, and administrations starting from a low base carry no legacy architecture that must be unwound. Connecting automatically to any external system still requires an accurate taxpayer register, so data integrity remains a prerequisite whichever stage the technology belongs to.
Not automatically. It depends on whether the process was transformed or only digitised. Where taxpayers have to extract information from their own digital records and re-enter it into a government portal, they are running two processes instead of one. The OECD identifies this as a risk of digitisation that stops short of transformation.
Footnotes
[1] Tax Administration 3.0: The Digital Transformation of Tax Administration — OECD
[2] Digital transformation of tax administration — OECD
[3] Tax Administration 3.0 and the Digital Identification of Taxpayers — OECD
[4] Tax Administration 3.0 and Connecting with Natural Systems — OECD
[5] Tax Administration Digitalisation and Digital Transformation Initiatives: Taxpayer touchpoints — OECD
[6] Tax Administration 3.0 and Electronic Invoicing: Initial Findings — OECD
