Client Estimate & Estimate Court

One real, already-settled target shared by the whole cohort, and a live cross-examination of the estimate you built to answer it

Due: 21 May 2027 · Weight: 55%

Nothing about this target is in the future. You are not forecasting anything — you are measuring something that already happened and was never counted directly.

The brief#

How many residential dwellings in the City of Melbourne were unoccupied for more than six months during calendar year 2026?

Every student answers the same question, for the same city, for the same year. 2026 has already finished by the time you read this. The number exists. Nobody publishes it.

Why this target#

Victoria’s Vacant Residential Land Tax applies to residential land vacant for more than six months in the preceding calendar year, and that half-year does not need to be continuous — the test is whether the property was actually lived in, not whether it was merely listed for sale or rent. Before the tax’s 2025 expansion statewide, it already applied to a defined set of inner and middle Melbourne councils that included the City of Melbourne itself, so this LGA has sat inside the tax’s scope for years before the one you are estimating.

You are not estimating tax liability, and this brief is not a tax exercise. You are estimating a count. You must not identify, imply, or classify any individual property or owner — that is a matter for the Commissioner of State Revenue, not for this assessment. Exemptions and individual enforcement decisions are entirely outside scope. What you are producing is a city-wide aggregate, built the way an analyst without direct access to SRO records would have to build one.

What you have, and what it is not#

The nearest published figure is the Census. On Census night 2021 the City of Melbourne recorded 26,072 unoccupied private dwellings out of 103,029. The ABS’s own dictionary is exact about what that means:

“An ‘Unoccupied’ private dwelling is most often a separate house, townhouse, apartment or flat, with no one staying there on Census Night… a form with one or more people present on Census Night is set to ‘occupied’; a form with all usual residents away on Census Night is set to ‘unoccupied’.”

That is one night, once every five years. The target you are estimating is a different kind of quantity entirely: not who happened to be away on one date, but who was absent, cumulatively, for more than half of an entire year. A dwelling can be empty on Census night and lived in every other week. A dwelling can hold a Census-night guest and stand empty the rest of the year. Neither the Census figure nor the six-month test measures the other, and nor does 2026 have a Census at all — 2021 is the nearest one you have, five years removed from your target year. This is Week 5’s bridge, built for real, against a real gap.

If you reach for older ABS data that once recorded the reason a dwelling was unoccupied — holiday home, for sale, temporarily absent — say so plainly and treat it as what it is: a much older collection method, not current evidence. Using it is optional, and it does not need to appear in every chain.

What Weeks 5 to 9 give you#

• Week 5 — the proxy-to-target bridge, and the discipline of naming its breaking condition before it matters.
• Week 6 — when sources disagree, find the axis they actually differ on rather than averaging them into a number about nothing.
• Week 7 — a real business case, and the accept/reject/widen discipline for defending a number under pressure. Treat this as the rehearsal for the defence you give in the Court.
• Week 8 — naming which of your inputs you share with everyone else answering the same brief.
• Week 9 — who is exposed if this number is too high, who is exposed if it is too low, and whether that is the same party either way.

What you produce, before the Court#

A written estimate: the decomposed chain; an assumption ledger ranked by how much each assumption moves the answer, not by how likely it feels; a stated range with a plain-language confidence; the proxy-to-target bridge with its breaking condition named; a note on any conflicting sources and which axis they diverge on; a declaration of the inputs you know you share with the rest of the cohort; and a stakes note naming who bears the cost in each direction of error.

Alongside it, a pre-hearing disclosure: the material facts and assumptions your estimate rests on, stated plainly enough that someone who has never seen your working could attack them without first having to guess what you meant. This mirrors what a real expert witness must provide — the Federal Court’s Expert Evidence Practice Note requires a report to state the material facts and assumptions each opinion is based on, so that any challenge to it can be aimed at something specific rather than at a guess.

Due 7 May 2027.

The Court#

Week 11 is not a rehearsal. It is where the estimate above is examined live, in front of the room.

You sit in a panel with other students who took the same brief. Your written estimate and disclosure sit in front of you — there is no requirement to work from memory. Every panel answers at least one question drawn from a common pool, so every hearing is measured against something comparable, and then faces live follow-up from the panel chair and from the other panellists.

In response to any challenge you may accept it, reject it, or widen your range — the same three moves Week 7 trained, live instead of on paper. Accept where the challenge brings new evidence, or exposes a flaw in your reasoning, that materially changes the assumption. Reject where it supplies neither. Widen where it reveals more uncertainty than your number implied without justifying a move in either direction.

An assumption that does not survive being questioned is not, by itself, a mark against you. A precise, well-justified concession is frequently stronger evidence of judgement than holding a position that should not have held. What is assessed is the quality of your response to the challenge, not whether your original figure emerged untouched.

After the Court#

A short revision note, applying the test Week 7 named: did something change, or only the room? Something changing covers all three of new evidence, a flaw in your reasoning that the question exposed, and uncertainty you had not recognised — a challenge can be decisive without a single external fact being new. Only the room changing means pressure, repetition, or the seniority of whoever asked. State which challenges moved your estimate and why, and which you held and why.

Declining to revise, where you can justify it, is a complete answer. This note is not marked on how much you changed — it is marked on whether your account of the change, or the refusal to change, actually applies the test.

Due 21 May 2027.

What is marked#

Written estimate and disclosure (60%). Whether the chain, the ledger, the range, the bridge, the source reconciliation and the stakes note are genuinely present and genuinely reasoned — not whether your final figure sits close to any particular number.

Estimate Court (20%). The quality of your live reasoning under challenge. A defensible concession is worth as much as a defensible hold.

Post-Court revision (20%). Whether your account of what changed, and why, actually separates evidence and reasoning from pressure.