Dubai publishes a decade of numeric targets. Almost no other city dares to be measured that way

A roadmap with numbers on it

The Dubai Economic Agenda, launched in January 2023 and known as D33, commits the emirate to AED 32 trillion in total economic targets by 2033, doubling the size of its economy and placing it among the top three global cities. It is structured as one hundred transformational projects. Cities publish visions constantly; almost none publish a decade of specific numbers against which they can later be measured.

Most civic strategy documents are written so that success is unfalsifiable: become a hub, foster innovation, attract talent. None of those can be scored. D33 does the opposite. It states where the numbers currently sit, where they should sit in ten years, and by when. That structure is common in corporate planning and rare in the government of a city, and it is the reason the agenda is worth reading as a product roadmap rather than a manifesto.

What D33 actually commits to

The specific commitments are unusually legible. Foreign trade in goods and services is to rise from AED 14.2 trillion over the previous decade to AED 25.6 trillion over the next. Foreign direct investment is to rise from an average of AED 32 billion a year to an average of AED 60 billion, reaching AED 650 billion in total across ten years. Digital transformation projects are to generate AED 100 billion annually. In parallel, the general budget cycle for 2026 to 2028 has been approved, tying the near-term fiscal plan to the longer agenda.

Why publishing a target is the risky part

Publishing a dated numeric target converts a political statement into something that can be shown to have failed.

The counterintuitive part is that publishing the numbers is the costly move rather than the promotional one. An unpublished internal target can be revised without anyone noticing. A published one creates a permanent public record and hands every future critic a scoreboard. Organisations avoid this precisely because it removes their ability to redefine success after the fact, and the willingness to accept that constraint is itself a signal about how confident the issuer is — or how little it fears being scored.

Targets are only real if they can fail

A target is only meaningful if there is a published baseline and a date, and both are present here.

Run the arithmetic and the targets become assessable rather than impressive. AED 32 trillion across ten years is AED 3.2 trillion a year. The foreign trade target is a 1.8-fold increase on the previous decade, and the FDI target is roughly 1.9-fold. The digital component, at AED 100 billion a year, is about AED 1 trillion over the period — roughly three per cent of the headline figure, which places it as a meaningful contributor rather than the centrepiece its prominence in the communications suggests.

Where a single-owner model breaks

A model with one decision-maker has no institutional mechanism for revising a target downward in public.

The structural weakness is not the ambition but the revision mechanism. In a system where budgets pass through a legislature, a target that is missed gets adjusted in public because an opposition forces the adjustment. Where authority sits with a single office, there is no institutional actor whose job is to demand a downward revision, and the likeliest outcome of a shortfall is quiet restatement rather than acknowledged miss. That does not make the targets false; it makes the reported progress against them harder to verify.

What the roadmap has produced so far

The near-term artefact is the budget. Approving a 2026 to 2028 cycle inside a 2033 agenda is the step that converts a decade-long statement into an allocation, and it is the part worth watching, because a roadmap that never reaches a budget line is a communications exercise. The relationship between the three-year fiscal plan and the ten-year targets is where the agenda either becomes operational or stays aspirational.

The strongest argument against this

The reasonable objection is that publishing targets costs nothing when no one can enforce them. A city that controls its own statistics, has no opposition to demand an accounting and faces no electorate can announce any number it likes, restate it later, and treat the announcement itself as the achievement. On that reading D33 is marketing with a spreadsheet attached.

That objection has real force and is weaker than it first appears, for a specific reason: the audience for these numbers is not domestic. It is foreign investors, multinational headquarters and sovereign counterparties, all of whom track the figures independently and none of whom are obliged to accept a restatement. A city whose entire economic model depends on inbound capital is unusually exposed to being caught missing a published number, because the people who would notice are precisely the people it needs.

The difference between a vision and a specification

Nearly every city publishes a strategy, and nearly all of them are written to be impossible to score. Becoming a hub for innovation, fostering a knowledge economy and attracting global talent are positions rather than targets, and no year-end review can find them unmet.

A specification looks different. It names the current value, the intended value, the unit and the deadline. Once those four elements are present, an outside party can independently calculate whether progress is on pace without any cooperation from the issuer.

That is the actual innovation in D33, and it is a governance one rather than an economic one. The targets may or may not be achieved; the decision to state them in a scorable form is the part other administrations could copy tomorrow and almost universally do not.

What the model is genuinely good at

A structure with a single decision-maker and no legislative cycle can do certain things that no democratic administration can attempt. It can commit to a ten-year plan without the plan becoming an election issue, approve a project in weeks rather than years, and reverse an internal decision without a public process.

This publication has made a version of this argument about industrial policy in other countries, and it holds here in a purer form. Speed and continuity are real capabilities, and much of what Dubai has built in two decades is downstream of them rather than of capital, which was available to plenty of places that did not build it.

The honest accounting is that the same absence of process which produces the speed also removes the checks that catch a bad decision early. Both effects come from one source and cannot be separated by preference.

The part of the record that a profile should not omit

In 2020 and 2021 the English High Court issued findings in proceedings concerning the ruler's family, including in relation to his daughters. The rulings are a matter of public record in a jurisdiction whose judgments are published in full.

This sits uncomfortably beside an analysis of economic targets, and leaving it out would make the piece a promotional document rather than a profile. A publication that reports a leader's stated goals while omitting adverse findings from a court that published them is not being neutral; it is selecting.

It is also relevant to the analytical frame rather than merely to the moral one. The central argument here is about accountability structures and what happens when a system has few of them. Findings issued by an external court are, in that specific sense, the exception that illustrates the rule.

A caveat on sourcing

The D33 figures used here come from the agenda as published by the Dubai government, which is the issuer of the targets and therefore not a neutral source for progress against them.

That is acceptable for describing what was committed to and inadequate for assessing achievement. Progress claims should be checked against independent trade and investment data rather than against subsequent government communications restating the same targets.

Readers should also be sceptical of circulating figures on Dubai's economy that carry no source at all. The published targets are precise and easy to verify; a great deal of what is written around them is not.

What transfers outside government

The transferable practice is narrow and demanding: state the baseline, the target, the unit and the date, then publish all four before you know whether you will hit them.

Most organisations resist this instinctively, and the resistance is rational — publishing a number you might miss is genuinely costly. What it buys is that everyone downstream can calculate the same answer you can, which removes an enormous amount of internal argument about whether things are going well.

The discipline that makes it work is the one nobody enjoys: a stated procedure for what happens when the target is missed, agreed in advance. Without it, a published target does not create accountability. It just creates a number that will eventually be quietly redefined.

How to read any published target

Check whether a stated target carries a baseline, a date and an owner before treating it as information.

Check whether the target names a baseline, because a goal without a starting point cannot be scored. Confirm there is a date. Identify who is accountable if it is missed, and if the answer is nobody, discount accordingly. Convert the headline figure into an annual rate, since decade-scale numbers are designed to sound larger than the yearly commitment they imply. Then look for the budget line, because the distance between a published target and an allocated one is where most strategies quietly end.

The first published shortfall is the real test

Every ten-year plan misses something in year three or four. What matters is what happens when it does: whether the number is restated with an explanation, quietly dropped from subsequent communications, or replaced by a differently defined metric. The handling of the first genuine shortfall will reveal more about the system than any progress report issued while the numbers are still on track.

Reassess against the annual trade and FDI figures rather than announcements, and compare each year's reported baseline against the one published in 2023, since silently rebased figures are the most common way a missed target disappears.

Editorial conclusion

Running a city against a published decade of numbers is a genuinely unusual act of governance, and the interesting part is not the ambition but the exposure. AED 32 trillion by 2033, with a stated baseline and a date, is a commitment that can be shown to have failed — which is what separates a target from a slogan. Whether the emirate hits it matters less, analytically, than whether it reports honestly when it does not. That is the number nobody has published yet.