Objections, steelmanned

2026-07-18 · wave 4 · OBJECTIONS lane. Rule: each objection gets its strongest form first; answers cite evidence (much of it via the Georgism wiki's verified pages) or concede. An objection we can't answer stays here marked open. Updated w11 (§8 corrected), w14 (§4 tightened), w30 (§§1, 3, 4, 7 refreshed against the transition package, the completed readiness map, the nine measured rolls, and NUMBERS.md — including one retired figure caught and replaced).

1. "You're confiscating my life savings" (the transition shock)

Steelman. Canadians were told for generations that home equity is retirement security. Announcing the settlement capitalizes the future charges into land prices immediately: above-share holders take a one-time wealth loss they could not have avoided by working harder or saving more — they bought at prices that embedded the old rules. This is the strongest objection to every land-rent proposal ever made, and Groundshare's equivalence to LVT+dividend imports it whole. See the wiki's treatment: the transition wealth shock.

Answer (partial — this objection is partly conceded; instruments specified w10). (a) The incidence data (analysis/household-incidence.md) shows the loss concentrates in the top wealth quintile plus entity holdings; the average owner household gains. (b) The drafted transition package (analysis/transition-path.md) is a 20-year phase-in plus a recent-buyer credit plus universal deferral: under it, announcement-day land prices reprice toward ≈37% of today's over the phase-in, and the wealth-conversion identity cuts the other way from the steelman — each person simultaneously gains a claim (≈$164k capitalized equal share), so most households get wealthier on announcement day, not poorer. (c) What cannot be honestly promised: that nobody above the threshold loses — recent leveraged buyers and large holders genuinely do, which is what the credit and deferral are for. The moral case must carry that weight openly — the current owners' gains were themselves a transfer from everyone else, capitalized. Floyd ratification pending on the w10 package (phase-in length, credit design).

2. "Grandma is asset-rich and cash-poor"

Steelman. A widow in a Vancouver house bought in 1975 holds $900k of land on a $30k pension. A ≈$22k/yr net charge (NUMBERS #13 crossover arithmetic) evicts her as surely as a market rent would. Evidence and honest treatment: the asset-rich/cash-poor objection.

Answer (solved in design, at a fiscal cost). Universal deferral: net charges on a principal residence may accrue against the estate (with interest at the government borrowing rate), settling at transfer. British Columbia has run exactly this machinery for seniors' property taxes since 1974 — it is boring, tested administration, not a novelty. Deferral converts the widow's charge from a cash problem into an inheritance reduction — which is where the moral logic says it belongs.

3. "The scheme eats its own base" (reflexivity)

Steelman. At r near the full rental yield, holding land above your share yields nothing; speculative demand dies and land prices fall toward capitalized use value — possibly a long way. The $7.9T pool the site advertises shrinks, the dividends shrink with it, and the promised ≈$9k/yr was an illusion measured at pre-announcement prices. See does LVT destroy its own tax base?

Answer (mostly answered — with a design consequence). The confusion is stock vs flow. What the settlement actually moves is the rent — the annual value of occupying ground — and rent is set by what locations are worth to use, not by the asset price. When the speculative premium deflates, the price falls but the rental value persists (that is the textbook result on the wiki page above). The honest statement: the $353–394B/yr flow estimate (net basis, NUMBERS #4) is anchored to rental value and survives price deflation; the $7.9T stock figure will genuinely shrink as speculation exits — and should, since that's the scheme working. Design consequence (adopted as working position): define the settlement on assessed annual rental value (T = s × (R̄ − R_held)), not on asset prices with a rate — the rate-on-price form is a presentation convenience, the rent form is the robust specification. Modeling the transition path stays open (ACADEMIC lane).

4. "Land can't be assessed accurately / assessment will be corrupted"

Steelman. Separating land from improvement value is contestable parcel by parcel; assessment machinery invites lobbying, and errors become transfers. Strong forms: land cannot be assessed · assessment corruption.

Answer (largely answered by existing practice — claim tightened w14, map completed w27, machinery demonstrated w29). Every Canadian province already runs mass assessment with appeals, and all ten provinces' publication practice is now field-checked (analysis/assessment-readiness.md): BC and Quebec publish the land split per parcel, Saskatchewan shows you yours on request, and the other seven publish totals only — while every one computes a land value inside its machinery. That the machinery works at scale is no longer an inference: we've measured nine full city rolls — 1.4 million parcels — from the published splits (data/land-shares.csv, analysis/bc-open-data-sweep.md). The gap is transparency, not capability, and it's a ready-made early ask: "publish the land values your assessors already compute." The wiki's pages document the accuracy literature (modern mass-appraisal methods hold land-value error within workable bounds) — and Groundshare's netting design lowers the stakes per household: most households' cheques change modestly with assessment error because charges and allowances offset. Residual risk (concentrated lobbying by large holders) is real and answered by transparency: publish the roll.

5. "People will game the attribution" (trusts, spreading, offshore)

Steelman. The scheme is person-based; wealth advisors exist. Land moves into trusts, numbered companies, or across family members' allowances; foreign owners hide behind chains. The wiki documents how hard beneficial ownership is: land ownership secrecy.

Answer (mostly structural). The design already breaks the main exploits: entities get no allowance, so incorporating land increases the charge — the gradient runs toward personal, transparent ownership, the reverse of most tax shelters. Spreading among family members uses allowances those people genuinely hold — that is the design working, not failing (a family of four pooling four shares is the intended protection). The genuine residual: attribution through trusts to persons, which needs the beneficial-ownership registries Canada is already building (federal CBCA registry live since 2024; BC's LOTA). Concession: administration here is real work and real cost; parcel-based LVT is simpler on exactly this margin.

6. "This is socialism / an attack on property"

Steelman. However framed, the state directs flows between private citizens based on what they own. Today a settlement, tomorrow the rate rises; property rights once conditional are never secure again. Strongest form on the wiki: is LVT socialism?

Answer. Groundshare strengthens the private property that liberal theory actually defends — the fruits of labor. Improvements, buildings, businesses, savings: untouched, unconditionally. What it conditions is the one asset nobody produced, on the oldest property-theory grounds in the classical-liberal canon (Locke's proviso; Smith's ground-rents "peculiarly suitable" for taxation; Paine; Friedman's "least bad tax"). The geolibertarian tradition exists precisely because this position is compatible with strong markets: geolibertarianism. And the zero-revenue design is the answer to the ratchet fear: the state keeps nothing — every dollar in is a dollar out to a citizen, visible on one schedule.

7. "Homeowners vote; this dies at the ballot box"

Steelman. ~two-thirds of Canadian households own; homeowners vote at higher rates and organize; every property-tax reform teaches the same lesson. The wiki's version: homevoters will block LVT.

Answer (the incidence table is the campaign). Under Groundshare the median homeowner gains (≈+$3,700/yr for the average owner household, NUMBERS #10; crossover ≈ $866k of property at average land share) — the homevoter coalition splits in a way plain LVT never achieves, because the dividend is fused into the same visible schedule as the charge. The nine measured city rolls localize the paying class precisely: scarcity-priced ground (Metro Vancouver's detached land at 73–85% shares, central Toronto) plus entities — while Quebec's metros, at one-third land shares, land overwhelmingly on the receiving side, homeowners included. The real fight is with the top-quintile and entity holders, who are few but loud. Open risk: loss-aversion beats expected gain in referenda; the narrative lane owns this.

8. "It kills the family farm"

Steelman. Farms are land-value-heavy by construction; a Saskatchewan grain family can hold $3M of farmland against modest income. The charge forces sales and consolidates farmland into corporate hands — the opposite of the intent. Wiki treatment: LVT hurts farmers.

Answer (CORRECTED w11 — conceded as a class, answered in design). The original text here claimed most family farms sit below the crossover. The measured data says otherwise: the average Canadian farm holds ≈$4.0M of land ($4.9M in SK/AB) — decisively above-share in every province (analysis/farm-cut.md). What carries the answer now is design, not distribution: (a) the rent basis — farmland rental yields run ~2–2.5%, not 5.5%, so the charge is tenant-rent-sized, not price-sized; (b) tenant parity — ~40% of farmland is already rented, and owner-operators would pay what tenant neighbours already pay, while farm households also receive their shares; (c) use-value assessment for farm-class land (every province's existing property-tax convention); (d) working-farm deferral to sale-out-of-farming or estate. Full exemption rejected (it recreates the farmland shelter). Farm organizations belong on the early-consultation list.

Status

Groundshare — a proposal in open development. Every number traces to a cited public source with its retrieval date; corrections are published, not erased. Rebuilt 2026-08-04 from the repo's research files.