Blog

Digging Deeper

Our own long-form writing on zoning law, land use practice, and what it takes to know what the code said on any given date.

Blog

What we’re writing.

Analysis from the ZoneLex team on the law changes reshaping land use practice, and the data problems underneath them.

Land Use PracticeTrent Livingston · September 2026

Finding the Law Was Never the Hard Part

Where land-use diligence actually breaks down, and what a decision-intelligence layer has to do about it. The code is online, the county runs a GIS portal, and the statute is one search away. Finding text is the easy part. The hard part starts the moment you have all of it in front of you.

A single parcel outline with city, county, state, federal overlay, comprehensive plan and decision record layers stacked above it and resolved into one aligned record — Six layers of authority, six publishing schedules, six definitions of current. One parcel.

Read moreRead less

Ask any land-use attorney what slows a matter down and you will rarely hear “I could not find the ordinance.” The code is online. The county runs a GIS portal. The state statute is one search away. Finding text is the easy part.

The hard part starts the moment you have all of it in front of you.

A system that was never designed to be read as one thing

Land-use professionals in the United States work inside a stack of authorities that was assembled over decades, by different bodies, for different reasons: municipal codes, county requirements, state statutes, federal overlays, comprehensive plans, zoning maps, agency guidance, and the public record of past decisions. Each layer publishes on its own schedule, in its own format, with its own idea of what “current” means.

A municipal code posted online may sit six months behind the last adopted ordinance. A staff report may cite a plan amendment that has not been codified yet. A state preemption statute may quietly override a local provision that still appears in the code as if nothing happened.

None of these sources is wrong on its own. The problem is that they do not agree with each other on effective date or level of authority, and nothing in the system tells you that.

This is also why pointing a generic AI model at these materials does not produce legal-grade output. The model can read. The failure is upstream. The underlying information lacks the standardization, the authority structure, and the contextual metadata a system would need to know that one paragraph controls and another has been superseded. Hand it a picture of a PDF and ask which rule governs a parcel, and it will answer confidently and often incorrectly. That is a data problem before it is a model problem, which is why we started there.

Land-use work is not a search problem

Even a perfect search engine only gets you to the starting line. The actual work looks like this:

Determine which authorities govern this parcel, and in what order.
Identify where those authorities conflict, and where one preempts another.
Separate what is permitted by right from what requires discretionary approval.
Understand the administrative and political record around comparable approvals: who objected, what conditions were attached, how the hearing examiner ruled the last three times.

That last item is where the real value sits. Regional builders and land-use lawyers keep telling us the same thing:

Finding the law is easy. Understanding the historical conversation and sentiment surrounding it is what wins approvals. Context over code.

And the pace of change is accelerating. State preemption bills, federal incentives tied to housing production, pattern books, single-stair reform: all of it is rewriting the relationships between layers of authority faster than most firms can track. Current-law awareness is table stakes. Historical awareness tells you how a jurisdiction got here. Pending-law awareness tells you what your client’s project will face at hearing next spring. You need all three, on every matter, at the same time.

Where a decision-intelligence layer fits

The market that serves this work is fragmented across code publishing, GIS data, zoning reports, feasibility tools, legal AI, and permit systems. Each one solves a slice. None of them connect a parcel to its full governing record and keep that record honest as the law moves.

That gap, the intersection of legal diligence, zoning intelligence, and dynamic regulatory monitoring, is the niche we are building into.

The approach starts with a parcel, because that is where every matter starts. From there: resolve the governing jurisdictions, assemble and version the applicable law, flag authority conflicts and pending changes, and connect the formal requirements to the staff reports, hearings, agendas, minutes, and prior decisions that explain how those requirements actually get applied. The output is a shared, source-linked record the whole team works from: attorney, planner, developer, and client alike, looking at the same binder.

What this should change on a matter

If a layer like this does its job, here is what changes, in the order it matters to the client:

Parcel-to-issue analysis gets faster.
Conflicts, preemption, and unresolved authority surface earlier, before design money is spent.
Preparation for agency discussions, negotiations, and hearings improves, because the record is already in hand.
Client exposure to delay, redesign, failed acquisition, or missed conditions goes down.
Awareness of legal change, current and historical, becomes continuous instead of a scramble.
One regulatory record replaces a dozen versions scattered across email threads.

Behind those sit the quieter wins that partners notice over a year rather than a week: more consistent work product, better delegation across the team, source-linked defensibility for every determination, easier client status reporting, institutional knowledge that stays with the firm when someone retires, and far less exposure to unsupported AI output.

What this is not

Land use is dealmaking. The attorney’s job is not to render a final verdict on a parcel. It is to navigate the client to the best achievable outcome under the code as it exists and the political reality as it stands. No platform replaces that judgment, and none should be issuing legal opinions.

What a platform can do is the digging. Assemble the governing record, show what changed and when, flag where authorities collide, and put the history of prior decisions in front of the practitioner before the hearing rather than after. That leaves the professional with more time for the part of the job that actually requires a professional.

The bottom line

ZoneLex is building the jurisdiction-aware decision-intelligence layer for U.S. land use, starting with land-use law firms and specialized legal teams, where applicability, provenance, and professional review carry the highest value. The same verified data infrastructure will extend to developer feasibility, in-house legal teams, and enterprise API licensing from there.

We do the digging. You do the dealmaking. Beta opens this November, and we are following the build in public along the way.

Join the beta cohort →

Follow the journey →

Regulatory & Legislative WatchZoneLex · August 2026

Florida Just Changed the Rules of the Zoning Fight

What HB 399 and SB 180 mean for land use practice, and why the paper trail is now the whole ballgame. Starting January 1, 2027, a Florida city cannot recommend denial without naming each specific incompatibility in writing, and any stricter land use rule adopted since August 2024 is voidable.

A city zoning code page stamped DENIED, with every objective criterion checked as met and only subjective reasons for denial left unchecked — Illustrative only. When the criteria are objective, a denial has to survive on paper.

Read moreRead less

On January 1, 2027, Florida’s HB 399 takes effect. It requires local governments to evaluate development applications against objective compatibility criteria. And here is the part worth rereading: before staff can recommend denial, they have to name each specific incompatibility in writing. If the applicant offers mitigation, the city has to issue written findings explaining why that mitigation does not work.

Running alongside it is SB 180, which voids any stricter land use rule adopted anywhere in Florida between August 1, 2024 and October 2027, and awards attorneys’ fees to whoever successfully challenges one.

Taken together, these two laws do not just tweak Florida’s zoning process. They change the nature of the fight itself. Here is how.

From politician to project manager

In a purely discretionary system, a land use attorney spends a lot of time lobbying. Meetings with council members. Coffee with the mayor. A project’s fate often came down to a popularity contest.

Under an objective standard regime like HB 399, the job shifts. Success relies less on winning the room and more on directing a team of civil engineers, traffic consultants, and planners to ensure the application flawlessly checks every objective box in the city’s code. The attorney becomes less politician, more project manager.

That is not a demotion. It is a different kind of dealmaking. Land use counsel has never been in the business of rendering verdicts. The job is navigating a client to the best achievable outcome under the code as it exists. Objective criteria regimes reward exactly that skill set, because the path to yes is now written down. The question is whether you can find it, document it, and prove you met it.

A stronger position when it goes to court

Historically, if a city council denied a project because residents complained it would ruin the neighborhood character, overturning that denial was an uphill climb. Courts defer to local governments on discretionary zoning decisions, and they always have.

HB 399 changes the math. If a city denies a project without producing the required written findings and specific areas of incompatibility, the attorney now has a much cleaner path to challenge that denial, typically through a writ of certiorari, or to force a settlement. The statute puts a burden on the city that did not exist before, and burdens that are not met become leverage.

Forcing the mitigation conversation

Because the law says cities cannot deny a project when mitigation measures adequately address the stated incompatibility, the attorney’s work becomes heavily focused on drafting legally binding mitigation agreements. Instead of accepting a flat no, counsel forces the city to the negotiating table: we will add the turn lane and the landscape buffer. Now prove, in writing, why that is not enough.

That is the dealmaker’s move. The law hands practitioners a mechanism for turning denial into negotiation, and negotiation is where good land use lawyers have always earned their keep.

The other side of the coin

None of this means cities are out of moves. Expect three countermeasures, and expect them quickly.

Hyper-specific objective criteria. If a city can no longer deny a project for being too dense, it will rewrite its Land Development Regulations to be restrictive by the numbers: strict 50-foot setbacks, 40 percent mandatory open space, exacting architectural design standards, stringent traffic concurrency metrics. A perfectly legal denial based on failing a mathematical standard rather than a subjective one.

Death by incompleteness. If a city cannot easily deny a project on its merits, procedural delay becomes the tool. Staff can repeatedly return an application as incomplete, requesting more engineering reports, more environmental studies, more traffic data. Time and money drain away, and many projects die before they ever reach an up-or-down vote.

Bulletproof staff reports. City attorneys will train planning staff to write highly detailed, defensively structured denial reports, backed by the city’s own outside experts, generating reams of technical data proving exactly why the applicant’s proposed mitigation fails.

The pattern across all three is the same: the fight moves from the podium to the paperwork.

The SB 180 problem nobody is talking about

Here is a question that sounds simple and is not: what did the code say, for this parcel, on this specific date?

SB 180 makes that question load-bearing. If any stricter rule adopted after August 1, 2024 is voidable, then every application, every denial, and every challenge in Florida for the next several years turns on a precise historical snapshot of the code. Which version was in effect when the application was filed? When was the amendment adopted? Does the stricter provision predate the window or fall inside it?

Anyone who has actually tried to answer that question knows the problem. Zoning codes live in PDFs, scattered amendments, and meeting minutes. Municipal websites often publish a current version with no reliable record of what came before. The gap between what the published code says and what the law actually was on a given date is not a hypothetical. It is where cases are won and lost, and it is exactly the kind of error that quietly sinks a vested rights argument.

In an objective criteria regime with a statutory voiding window, the side with the better historical record has a structural advantage. Full stop.

The bottom line

Laws like HB 399 genuinely tip the scales toward applicants and their counsel. They take away the power of the angry mob at the council meeting, because a council can no longer vote no simply because a hundred neighbors showed up in matching red shirts.

But they also start an arms race. Land use attorneys must become deeply proficient in technical, administrative practice, and cities will respond by making their objective codes as dense and difficult to navigate as legally permissible. The winners will be the practitioners who can track every amendment, prove what the code said on any given date, and check every box before the city ever gets the chance to say it was missed.

That is the unglamorous middle of this story: the data work, the version tracking, the paper trail. It is the problem we spend every day at ZoneLex working on, because Florida is not an outlier. It is a preview.

Sources:

Bill Summary, CS/CS/CS/HB 399 (2026), The Florida Senate →

Bill Summary, CS/CS/SB 180 (2025), The Florida Senate →

Florida’s HB 399 Creates New Tools to Move Real Estate Projects Forward, Holland & Knight →

Florida Legislature Retroactively Prohibits More Restrictive Regulations on Development, Weiss Serota Helfman Cole + Bierman →

AI & Legal ResearchZoneLex · August 2026

Lessons Learned: What the Hallucination Crisis Actually Taught Us About Legal AI Architecture

Hallucination is a real risk in legal AI, and it turned out to be the smaller half of the problem. The bigger lesson was architectural: zoning law has a lifecycle, and an answer that is right on Tuesday can be wrong by Thursday.

Diagram of an evidence lifecycle loop: source, capture, verify, answer, re-check, then back to source — Verification is a stage in a loop, not a box you tick once.

Read moreRead less

When the first wave of legal AI hit the profession, hallucination got all the attention, and for good reason. A fabricated citation is easy to spot, easy to sanction, and easy to write about. Two years on, our own view has shifted. Hallucination is a risk we design against every day, but it was never the mountain we thought it was at the start.

The quieter failure is staleness

In land use, the source material itself moves. A council adopts an overlay on a Tuesday night. A planning department reposts a code section with a silent amendment. A condition of approval travels into a staff report and never makes it back onto the zoning map. None of that is a hallucination. Every word of it is real. It is simply no longer current, and a model repeating it confidently is wrong in a way that no citation check will ever catch.

The lesson is architectural, not moral

Most of the industry response to hallucination has been about restraint: tighter prompting, more disclaimers, a human somewhere in the loop. All useful, and all built on the same assumption, which is that the unit of work is the answer. We came out of this believing the unit of work is the record. If the record is sourced, dated, and watched, the answer takes care of itself. If it is not, no amount of caution at the output stage saves you.

Zoning law has a lifecycle, so the answer needs one too

That is why we stopped drawing our architecture as a pipeline and started drawing it as a loop. Source, capture, verify, answer, re-check. The re-check is the part that most legal AI conversations skip, and it is the part land-use professionals actually live with, because the ordinance that governed the parcel at diligence may not be the ordinance that governs it at entitlement.

Three things we took from it

Provenance before fluency. A short answer that points at the ordinance beats a beautiful paragraph that points at nothing. If a fact cannot be traced to a public record, it does not belong in the workspace.

Dates are facts. Every captured provision carries an as-of date and the version it came from. “Current” is a claim, and claims need evidence.

An answer is a subscription, not a receipt. If the underlying source changes, the people who relied on it should hear about it. That is a product decision, but it started as a lesson about what the hallucination debate was missing.

Hallucination is a design constraint. Staleness is the business problem.

None of this removes the professional from the work, and it is not meant to. Land use is deal work as much as legal work. The value an attorney adds is rarely being provably right on the record. It is knowing the law well enough to steer a client toward the best available outcome. Our job is to make sure the record they are steering from is complete, sourced, and current.

Housing PolicyZoneLex · August 2026

The Missing Middle Did Not Disappear. It Was Zoned Out.

Street elevation with single-family homes, ghosted outlines of a duplex, triplex, fourplex and townhome row labeled zoned out, and mid-rise apartments — Most residential land in America allows a house or a tower, and almost nothing in between.

Duplexes, triplexes and fourplexes used to be ordinary housing. Today they are illegal to build across most of the residential land in the country, and the reason is written in the zoning code, not the market.

Four panels: a fourplex, the lot consumed by surface parking, a denied underground garage and third story, and an empty lot — The parking math that kills the fourplex. Every compliant answer either eats the lot or gets denied.

Read moreRead less

Ask most people to picture missing middle housing and they cannot, which is the whole problem. Duplexes, triplexes, fourplexes, courtyard apartments, small cottage clusters: buildings that hold more than one household but read like a house from the sidewalk. They used to be ordinary. Walk any pre-war streetcar neighborhood and you are surrounded by them.

It did not die of market forces

The missing middle did not lose a competition. It was legislated out of existence, district by district, over about seventy years. Single-family-only zoning removed the use. Minimum lot sizes removed the arithmetic. Setbacks, lot coverage caps, and floor area ratios removed the envelope. What is left on most residential land in America is a detached house or, in a narrow band somewhere else in town, a mid-rise. Almost nothing in between is legal to build by right.

Parking is the quiet one

Density caps get the headlines, but parking minimums are often what actually kills the building. Require a spot and a half per unit and a fourplex owes the city six stalls. On a small lot, surface parking eats the yard and then the building. Go structured and you are paying tens of thousands of dollars per stall. Go below grade and you have added excavation, shoring, waterproofing, ventilation, drainage, and a structural engineer to a project with four doors to spread the cost across. The math never closes. The code never had to say “no”—it just made yes unaffordable.

Why environmental and transit advocates are in this fight

This is where land use stops being a housing story. Up-zoning near transit, removing or capping parking requirements, and legalizing modest density are the levers that make walkable neighborhoods possible at all. Fewer required stalls means less impervious surface, less stormwater to manage, less heat island, fewer vehicle miles travelled, and a transit line with enough riders within walking distance to justify running it. A lot of the attorneys doing zoning work are environmental practitioners, and this is why: the ordinance is the instrument.

The unglamorous middle

There is a version of this problem in our own work. Everybody wants to talk about the model. The value is in the unglamorous middle: reading the PDFs, normalizing the definitions, reconciling the map to the text, keeping the dates straight. Reform only becomes real when somebody can say precisely what a parcel allows today. Right now that answer takes weeks, and reform moves at the speed of that answer.

The missing middle is not a design problem wearing a permitting costume. It is a permitting problem wearing a design costume.

If you have seen a local example of this—a fourplex that took two years, a parking variance that decided a project, an infill push that stalled—we would like to hear it. That is the raw material for everything we build.

Housing PolicyZoneLex · August 2026

The Closing Cost Trap: How First-Time Buyers Get Squeezed Out Long Before They See a Zoning Map

An FHA buyer who needs the seller to cover closing costs has to offer enough to make it worth the seller’s while, and then the appraisal has to support it. In a hot market that circle almost never closes.

Two bars comparing an offer price under the appraised value line with the same offer plus seller paid closing costs pushing above it — When concessions get financed into the price, the appraisal is where the deal quietly dies.

Read moreRead less

There is a small mechanical detail buried in first-time homebuying that decides more outcomes than almost anything in the zoning code, and almost nobody outside the transaction ever sees it.

The mechanics

An FHA buyer is, by design, someone with a small down payment. Getting to the down payment often uses up the cash. Closing costs—origination, title, escrow, prepaids, recording—are a separate several thousand dollars that has to show up on the same day. The standard fix is a seller concession: the seller agrees to cover the buyer’s closing costs out of the proceeds.

Sellers do not do that for free. To get one to agree, a buyer generally has to come in at or above asking, so the concession is effectively financed into the price. Which brings in the appraisal. The loan is sized against appraised value, not against the contract. If the appraisal does not support the higher number, the lender will not lend the difference, and the buyer has to bring it in cash—the exact cash they did not have in the first place.

In a soft market this works. In a hot one it does not.

When inventory sits, sellers pay closing costs and appraisals come in fine. When there are six offers by Sunday, a seller has no reason to take the offer that costs them money, and the comps are already stretched. The FHA buyer is not outbid on price. They are outbid on structure. The circle simply never closes.

Why a zoning company is writing about this

Because the condition that makes it impossible is supply, and supply is a land-use output. The entry-level home that would clear this hurdle is exactly the product that got zoned out: the small-lot house, the duplex half, the townhouse, the fourplex unit. It is not being built because in most places it cannot be built, and the ordinance is the reason.

Follow that thread and it runs everywhere: household wealth formation, commute lengths and traffic, transit ridership, stormwater and impervious surface, school enrollment, municipal fiscal health. The why behind fixing land use is not one story. It is a dozen of them sharing a root cause.

The first-time buyer is not priced out at the closing table. They are priced out at the zoning hearing, years earlier.

This is descriptive, not advice: the point is simply that a financing rule and a zoning rule are quietly deciding the same outcome, and only one of them ever gets debated in public.