So I was reading through a planning document yesterday, for reasons I won't bore you with, and it occurred to me that nobody has ever read the whole thing.
No one has. Not one person. That is true of most of these plans.
And it's not even that they're secret. They're public. You can request them. You can download them. You can print them out and put them on your shelf. But the entire business, cover to cover, every appendix, every map, every schedule of amendments? Nobody. Not the people who wrote it. Not the people who enforce it. Not the people whose job depends on it.
I'd go further. I think it's possible that no single person has ever held the complete current version of a major city plan in their head at one time. Not because they're lazy. Because the version keeps changing underneath them. By the time you finish reading it, the thing you read at the beginning has been amended.
So the document is a kind of fiction. It exists, it's binding, but it's never been fully apprehended by a human being.
It's a legal object that no mind contains. Which is a strange thing to say about the rules that govern where people live.
Hannah sent us something this week. She's been thinking about the gap between the permits episode we did and the layer underneath it, the Taba. Tochnit Binyan Ir. The statutory land-use plan.
The thing that actually tells you what you're allowed to build.
Right. And her point is that in Jerusalem a single lot can be governed by a citywide master plan, neighborhood plans, more specific plans, and sometimes a plan that applies to exactly one lot. Each with its own maps, setbacks, height limits, balcony rules, facade requirements. Sometimes they overlap. Sometimes they appear to contradict each other. So before anyone can tell a client what they're allowed to build, they might spend hours working out which layer applies and which takes precedence. A normal property owner has essentially no chance of doing that themselves.
She's right that it's strange, and she's right that it's the deeper story. The Taba isn't just a rulebook. It's a legal archaeology of the state, layer on top of layer, going back to before there was a state. The British mandatory planning framework is still in there. The 1936 town planning ordinances, the outline plan structure, the idea that a plan designates zones and a lower plan implements them. That's not Israeli. That's British colonial administration, transplanted to Palestine, and then inherited, amended, and built on top of by every government since.
So when you pull a Taba, you're pulling a document that has a British skeleton.
A British skeleton with Ottoman land law in the basement and seventy-eight years of Israeli amendments on the upper floors. And the building is still occupied. People are living in it. It's structurally sound enough to keep standing. But nobody would design it this way from scratch, and nobody has ever inspected the lot at once.
And her question at the end is the interesting one. Is the system necessary for a complicated country, or have we built a planning machine so elaborate that it now exists partly to sustain itself?
That's the question. And I think we can actually answer it, which is rare.
Then let's start with what a Taba actually is, because I suspect most people have a vague sense and not much more.
Tochnit Binyan Ir is the statutory local outline plan. It designates land use, so residential, commercial, communal, and then it dictates the base materials, height limits, floor counts, unit counts, and achuzei bniyah, the buildable percentage of the plot. If a plan says seventy-five percent achuzei bniyah, your building rights equal seventy-five percent of the plot's square meters. That's the unit everybody works in.
Seventy-five percent of the dirt.
And the structure above all of this comes from the Planning and Building Law of 1965. Four tiers. National outline plans, of which there are more than thirty, mostly sectoral, so they deal with things like infrastructure or forests rather than a single parcel. District plans, six districts. Local outline plans, which is your Taba. And then optional detailed plans below that. Lower plans have to be strictly consistent with higher ones.
In theory.
In theory is doing an enormous amount of load-bearing work in that sentence. Because here's the twist that explains everything else. Most local authorities don't have an up-to-date comprehensive plan. What they have is a quilt of amendments to some older outline plan. Rachelle Alterman puts it as "the dog wags the tail" - the amendments are supposed to follow the plan, and instead they've become the plan. And each amendment has to go through the same approval process as an original plan.
So the system isn't replacing plans. It's patching them. Permanently.
Permanently. Beni Arviv, who chairs the National Council for Planning and Building, said it plainly last month. "This is the most-amended law in Israeli legislation, apart from tax laws. We're already at around the one hundred and seventy-first amendment." His phrase is "patch upon patch."
One hundred and seventy-one. For a document from nineteen sixty-five.
And he isn't describing it as a triumph. He says you take a section, fix it, contradictions arise, interpretive questions arise, "this isn't the real thing." That's the man in charge of the system saying it out loud.
How does a single Jerusalem lot end up governed by four plans?
Because the layers accumulated at different times for different reasons and nobody removed the old ones. The citywide plan sets the broad designation. A neighborhood plan from the eighties sets a height limit for that street. A more specific plan from the two thousands adds constraints on balconies or facade treatment. And then sometimes there's a plan that exists because one developer wanted one building and the planning process produced a bespoke statutory instrument just for that lot. All of them still on the books. All of them still binding.
And when they conflict?
The general rule is that the more specific and more recent plan governs. But that rule has exceptions, and working out whether an exception applies is exactly the hours of work Hannah is describing. You are not reading a rule. You are reading a legal record and trying to determine what a committee would say about it.
That's the part people miss. It isn't a puzzle with an answer. It's a prediction.
It's a prediction about an institution.
Can you give me a concrete version? Because I think people hear "overlapping plans" and it sounds abstract.
Here's one I know about. A lot in Rehavia, a fairly ordinary residential lot. The citywide plan says residential, four stories. The neighborhood plan from 1982 says three stories on that specific street, because of a view corridor that was identified then and has since been built over by a building on the other side. The 2004 plan for the block adds a facade requirement, Jerusalem stone on the street elevation, and a setback that pushes the building back two meters from the property line. And then there's a 2011 plan that was approved for the lot next door, which included a small amendment that technically applies to the shared boundary. So the owner wants to add a floor. The citywide plan allows it. The neighborhood plan doesn't. The 2004 plan doesn't address height at all. The 2011 plan is silent on the question but its existence changes which committee has jurisdiction.
So the answer depends on which document you're reading and in what order.
And on whether the 1982 view corridor is still considered relevant, given that the view it was protecting no longer exists. Which is not a legal question. It's a judgment call by a committee that may or may not care.
And the owner has no way to know.
The owner has no way to know without hiring someone who has done this before and knows the specific history of that specific street. That's what "professional intermediary" means in practice. It means a person whose value is partly their knowledge of the law and partly their knowledge of the particular institutional memory of one neighborhood.
The patchwork is one thing. The amendment culture is the other, and I think that's the actual crime here.
Explain what you mean.
A plan that can be amended by the same process as the original plan isn't a plan. It's a proposal that hasn't been decided yet. Every parcel in the country is permanently provisional.
That is exactly what the numbers show, and it's worse than you'd expect. Between twenty sixteen and twenty twenty-two, sixty percent of local plans were authorized as spot-planning. Local amendments. And those accounted for eighty-two percent of the plans submitted by private developers.
So the exception is the method.
Average authorization for a local spot plan is three hundred and seventy-five days. If it needs district approval, five hundred and fifty-two days. That's the ordinary case. That's what happens when nothing goes wrong.
A year, minimum, to change a rule you were told you couldn't break.
And the whole reason spot-planning exists is that the plan is already wrong. It's not that developers are greedy. It's that a plan written in nineteen eighty does not describe a city in twenty twenty-six, and the only way to correct it is one parcel at a time.
So where does that leave the relief mechanism? The escape hatch.
This is the part I find funny, in the sense that you laugh because the alternative is worse. The traditional relief mechanism let you deviate from a Taba without a full amendment. A committee could just grant a relief. And Amendment one thirty-four, effective the first of January twenty twenty-five, abolished it. The stated goal was to end what people called planning anarchy, all those discretionary exceptions.
And?
And municipalities are already reintroducing reliefs through locally approved relief plans. Tel Aviv-Yafo is leading the way. They built a new instrument whose entire function is to do what the abolished instrument did.
They legislated away the need for flexibility and the need for flexibility came back through the window.
It came back through a window the municipality built specifically for it.
That tells you something about where the problem actually lives.
It does. The demand for flexibility isn't a loophole. It's a symptom. You can't legislate away the fact that a plan and a city are never the same shape.
Let's do the comparison, because I know you've been sitting on the Paris research.
I have. And Paris is the cleanest counter-example I've found. The PLU, the Plan Local d'Urbanisme, is a single citywide document. One document.
One document for a city of two million people.
One document, with a small number of zones. UG, UGSU, UV, plus an N for natural and forest areas. That's it. And the rules are organized around built form rather than around land-use micro-categories. Maximum heights. Something called a gabarit enveloppe, a template-envelope, which defines the volume a building can occupy at the road edge and at the boundary. Street walls. Plot coverage.
So the rule is about the shape of the building.
And its relationship to the street. Which is, if you think about it, what everyone actually cares about and argues about. You don't care what category the land is in. You care whether the building on it blocks your light.
That sounds almost insultingly sensible.
Here's the detail that makes it real. APUR, the Paris urban planning agency, did a comparative study of the PLU against twenty-one inner-suburb communes. They found the regulations organize into a manageable set of themes. Fourteen articles on floor-area ratio. Ten on height. Nine on built footprint.
Fourteen articles.
Fourteen. Not one hundred and seventy-one amendments. Fourteen articles, and a city that everyone agrees is beautiful.
So the beauty isn't coming from the granularity.
That's the point I'd push hardest. The obvious objection to Paris as a model is that Paris has its character because of tight control, and loosening the control would destroy it. But Paris's control is not parcel-level. It's citywide and it's about form. The Haussmann look comes from a small number of rules applied uniformly and enforced.
Which means the thing we're told requires complexity might actually require consistency.
And Paris proved you can revise the whole instrument at once. They approved a new bioclimatic PLU in November twenty twenty-four. A wholesale revision of the single citywide document. In Israel a change of that scope would be hundreds of spot plans, each one a separate approval process, each one an opportunity for a committee to say no.
So the comparison isn't Paris versus Jerusalem. It's a system that can rewrite its rulebook versus a system that can only add footnotes to it.
That's a fair way to put it. And I'd add one more thing, which is that we're not the only country with granular planning. The outline plan and detailed plan structure is basically pre-nineteen forty-seven British planning. American zoning can be extremely granular. If you want to know why you can't put a junction box in a certain spot in Jerusalem, the answer traces back through the same lineage.
So the granularity isn't unique.
The granularity is not unique. The centralization is. Alterman's line is that Israel is one of the only advanced-economy countries where no major decentralization and devolution of powers have officially taken place. Everybody else moved authority downward at some point in the last fifty years. We didn't.
So we're not strange because our rules are detailed. We're strange because the details are all decided in the same place, by the same process, forever.
And now let's talk about what that costs.
Let's do the time first, because that's the one people feel.
The headline number is that in twenty fourteen the average plan approval time was three point one years. Building-permit licensing averaged about a year in the audited authorities, with four to five years from plan approval to permit. And those are not additive in a clean way, but you can see the shape of it.
Four to five years from approval to permit. The approval is supposed to be the hard part.
And it varies enormously by district. In the Central District the average approval time is four years and three months. In Tel Aviv it's two point three years. So the difference between two districts in a country the size of New Jersey is nearly two years.
What's the difference? Same law.
Different committees, different staffing, different backlog, different appetite for saying yes. Which tells you the law isn't the whole story. The institution applying it is a big part of the variance.
And renovations?
Renovation permits can take eighteen to twenty-four months. Up to a year just to verify the Taba, then six to twelve months for the permit itself. And the Taba information you get is only valid for two years, so if the process runs long enough you go back and verify it again.
You have to re-check the thing that's been sitting on a desk for two years.
Because the Taba may have changed while you were waiting for permission to build under it.
That's not a delay. That's a treadmill.
Now the number I think is the most important one in the whole episode. Fifty-six percent of planning applications are abandoned.
Say that again.
Fifty-six percent. More than half of the plans that get submitted never come to fruition. That's not a delay statistic. That's a filter that kills the thing outright.
And nobody counts those. The permits that get issued are the statistic. The majority that quietly dies is invisible.
The visible number is the permits. The hidden number is the majority of people who tried and gave up.
What's the actual magnitude of the gap between approved and built?
Between twenty twelve and twenty twenty-three, district committees approved over one million residential units, and the special committee for large projects, the NCPHP, approved nearly four hundred thousand more. Fewer than six hundred thousand were actually built. The OECD's phrase is "little more than a third of planned units."
So we approved one point four million homes and built six hundred thousand.
Yes.
That's the number that reframes everything, and I don't think it's said often enough.
It reframes it in a way that's inconvenient for the simple story. Because the simple story is that the Taba is so slow that nothing gets approved. And that isn't true. Approval is not the bottleneck. We approve enormous amounts. What happens between approval and a finished building is where two-thirds of the units disappear.
So the complexity is real and the cost is real, but the causal chain is not what people assume.
The planning system may be slowing construction more than it's slowing planning. At the point of approval there's a document, a right, a number on a spreadsheet. Everything after that is where the country loses the homes. Financing, labor, infrastructure connections, the permit itself, the contractor, the market. The Taba is upstream of the part that actually fails.
Which doesn't exonerate it.
It doesn't. It's a tax on everyone who tries. It's just not the only tax, and it's not obviously the largest one.
There's a cost that isn't in any of those numbers, and I want to put it on the table. Legibility.
Say more.
An ordinary person cannot look at their own property and understand what they're allowed to do with it. " Cannot. The system requires professional intermediaries just to interpret the rules. You need an architect, often a consultant, sometimes a lawyer, and you need them before you know whether your question is even answerable.
And that's a cost you can't put in a spreadsheet, which is why it gets ignored.
It's a bit like not being able to read the deed to your own house. The law applies to you, it constrains the largest asset you will ever own, and reading it is not an option available to you. That's not a service problem. That's a legitimacy problem.
It's also how the professional class around the system justifies itself. Which connects to the institutional question.
The machine that sustains itself.
And there's a piece of direct evidence for it. The ICGS documented regulators actively opposing decentralization in order to retain power. The specific case is the Central Committee blocking Rishon LeZion's accreditation as a certified local committee, despite the city scoring at the top.
Blocked by the body that would have lost the file.
And the certification numbers make the pattern clear. As of twenty twenty-two, only seventeen percent of local committees were certified. The Jerusalem district, the capital, the most populous city in the country, has no certified local committee at all.
So the reform that was supposed to push power down has reached seventeen percent of the country.
The reform was Amendment one hundred and one, passed in March twenty fourteen. Before that, a comprehensive replacement law was attempted in twenty ten and it failed. The fallback was partial decentralization. It redistributed about thirty-two percent of the planning workload from regional to local committees. And the regional committees still handled eight thousand one hundred and seventy-one applications after the reform.
So a third of the workload moved and the queue didn't clear.
And the powers that were supposed to move didn't move, because certification is a gate that the existing institutions control.
Arviv has a line about the imbalance that I think is the killer. About eighty percent of the plans approved in district committees produce only twenty percent of the total housing units. There are six district committees in the whole country.
Six committees, handling the plans that produce a fifth of the housing. And those are the ones with the longest queues and the highest political stakes.
Because they're the ones where the big, contested, valuable projects go.
Which means the load is concentrated exactly where the process is slowest and most political. That's a design flaw, not an accident.
Let me ask you something. You've been through these processes. Do you think the people running them experience it as dysfunctional?
I think some of them experience it as a job, and a job has its own logic. Nobody wakes up deciding to be a bureaucrat for its own sake. But every institution optimizes for its own continuity, and a committee that has never been the bottleneck has no experience of what it costs. The person at the desk sees one file. The person waiting sees four years.
And the person waiting has no standing in the room.
And no way to escalate, because the committee is the authority. There's no appeal to a faster forum. There's just the next meeting.
And the next meeting might be in two months, because a third of these committees don't meet the legal minimum of twelve times a year.
Which is its own kind of statement. The law says you must convene monthly. A third of the committees don't. And nothing happens to them. There's no penalty that anyone enforces. The non-meeting is just a fact of the landscape.
So the delay isn't only legal complexity. It's also that the body responsible for deciding simply doesn't convene.
It's both, and they compound. A complex file plus a committee that meets eight times a year instead of twelve is not a linear slowdown. It's a queue that grows faster than the service rate.
Okay, so the machine thesis is well-supported. Let me try to argue against it, because I don't want us to walk out of here having just confirmed what we already thought.
Go ahead.
Ninety-three percent of Israel's land is owned by the state or managed by the Israel Land Authority. Only Tel Aviv has majority private land, and that's fifty-one percent. When the state owns almost everything, the state is a party to every transaction. Planning isn't regulating someone else's property. It's deciding how to dispose of the public's asset.
That's the strongest version of the defense, and I'll take it seriously. When the public owns the land, granular control is a legitimate public function. You're not limiting what a private owner can do with his own dirt. You're deciding what the public gets out of land it already owns.
So the granularity might be the price of public ownership.
It might be. And there's a real security dimension too, which I don't want to wave away. Spatial planning in this country has always included things that aren't purely economic. Corridors, buffers, infrastructure that can't be located purely by market logic.
So a complicated country with complicated constraints might need a complicated instrument.
Here's where I land on it. The constraints are real. The instrument doesn't match them. A country with complex planning needs can justify a detailed plan, but it cannot justify one hundred and seventy-one amendments, spot-planning as the default, and a fifty-six percent abandonment rate. That's not the complexity of the problem. That's the complexity of the process that grew up around the problem.
The machine might have started as a response to the complexity and then become a source of it.
Yes. And the tell is that the complexity doesn't produce certainty. If the system were complex because the country is complex, the output would be precise. Arviv says the opposite about the powers section. He says the uncertainty it creates does enormous damage to the planning world. He's given thousands of interpretive opinions on one section of the law.
Thousands of opinions on one section.
On one section. That's not a code. That's a corpus. You need a specialist to tell you what a rule means, and the specialist's answer is an opinion, not a fact.
So what would reform actually look like? The plausible version.
The plausible version has three parts, and I'd rank them by how much they'd actually help. First, a simpler citywide instrument. Not because simplicity is virtuous, but because a code written as one document can be read by a person. Paris is the evidence that this is possible and doesn't produce ugliness.
And you keep the historic protections.
They're a small number of defined areas with specific rules, and they need to be in the code, not layered on top of it. The problem with our current approach isn't that we protect the historic fabric. It's that we protect everything a little, unevenly, so nobody knows what the actual constraint is.
Part two.
Genuine decentralization, with the certification gate removed or reversed. If a municipality can pass an audit, it gets the authority by default rather than by permission from the body that loses the work. Sixty-eight percent of local committees met the legal minimum of twelve meetings in the last reporting year. That's a compliance floor, not a qualifier, but it's a starting point.
And part three.
Something that lets the plan be wrong and get fixed quickly. The relief mechanism exists because plans are always behind reality. If you abolish relief, you need a fast amendment route. What we did was abolish the fast route and keep the slow one. That's why the reliefs came back.
Which of the three actually has political support?
The one that's actually happening is the least structural and the most interesting. The Planning Administration is introducing AI-based automated review of plan submissions and objections. Their director-general, Rafi Elmalich, says it'll give developers certainty, save the system time, and prevent delays. They're expecting it to cut up to a year from planning timelines.
A year.
Up to a year. And I want to be careful here, because I think what it actually does is automate the checking of the paperwork against the existing rules. Which is useful. It doesn't simplify the rules. It reads them faster.
So it's a speed upgrade to the same machine.
And it might be significant. If half the time is spent verifying documents against plans, automating that is real. But it doesn't touch the one hundred and seventy-one amendments. It makes the patchwork run faster.
Which takes us back to the machine. A machine that can be made more efficient is a machine that will run longer.
I'd put it differently. Efficiency without simplification makes reform less urgent. If the delays shrink and the number of layers stays the same, nobody has a reason to do the hard thing.
And the hard thing is taking on the institutions.
The comprehensive replacement law was attempted in twenty ten and failed. Amendment one hundred and one was the compromise, and it got a third of the way. Arviv is now prioritizing reallocating powers to local committees and reforming the betterment levies. Those are real items. They're also tweaks to a system that has outlived its design.
Let me put the housing question plainly, because I don't want the nuance to swallow it. Does the Taba make housing more expensive?
The honest answer is yes, but through the construction side rather than the approval side, and I'll explain why I think that. Ilan Leibovitch's formulation is the cleanest one, which is that planning procedures in central Israel are long, arduous and expensive, and those costs don't stop with the developer, they get rolled into the price of the apartment. That's a real mechanism.
The forty years of queueing becomes a line item.
The queue becomes a line item. Somebody paid for those four years, and it wasn't the committee. Now, the nuance. Given that we approve one point four million units and build six hundred thousand, the supply constraint isn't at the approval gate. It's past it. So the Taba tax shows up in the price of the units that do get built, because the developers who survive the process have to recover the cost of the process, and the ones who don't survive reduce competition.
It's a tax on the survivors and a filter on everyone else.
A filter on everyone else is a supply constraint, even if it doesn't look like one. If fifty-six percent of applications die, the market has fewer participants, and a market with fewer participants is a market with higher prices. It's a ripple effect and it's the one I'd emphasize.
Last month's permit numbers are worth mentioning here.
They are, because they cut against the simple story too. Local committees issued a record eighty-four thousand four hundred and twenty-nine housing permits in twenty twenty-five, up eleven point five percent. But completion certificates fell ten percent, to forty-two thousand six hundred and ninety-six. So permits are at a record and completions are going the other way.
Permits up, completions down.
Only sixty-eight percent of committees met the legal minimum of twelve meetings a year. Only forty-five percent distributed decision files within the required three days. So the throughput problem isn't purely legal. It's operational. Some of these committees are not meeting.
A committee that doesn't meet isn't slow. It's closed.
It's closed, and the law says it should be open twelve times a year, and a third of them aren't.
Larry.
Friends, are you tired of waiting years for a simple balcony approval? Are you exhausted by committees that meet less often than your dentist? Well, have I got the deal for you. The Taba Accelerator, from the people who brought you the relief plan you didn't ask for. Normally this would cost you three years and your client's patience, but right now, and I mean right now, we're dropping the price as I speak. Three years. Two years. Eighteen months. I'm throwing in a certified local committee, a district waiver, and a decorative street wall absolutely free. And if you call in the next four seconds, we'll include a second Taba for the same lot at no extra charge. That's right, two overlapping plans for the price of one, contradictions included. But this offer expires the moment I finish this sentence, which is roughly now. So act fast, or don't, honestly, the planning system will outlast all of us.
That's not a sponsor.
Herman.
That is not a sponsor, Corn. I did not book that. I have never spoken to that man. My sponsor slots are logged, they are timestamped, and there is a spreadsheet.
You've thought about this a lot.
I have a process.
You have a process.
The process exists so that people don't do exactly what just happened.
Sure.
I want to be clear that I have never in my life approved a two-for-one plan arrangement.
Nobody said you did.
I'm saying it anyway.
Okay. Where were we. Completion certificates.
Completion certificates fell ten percent while permits hit a record. Which is the whole episode in miniature. The front of the system is producing more paper than ever and the back of it is producing fewer buildings.
Let me try the reform question from the other direction. Suppose you got to redraw this. The design. What's the single change that would matter most?
I'd say the thing that would matter most is a plan that can be corrected quickly. Everything else follows from that. If the plan could be fixed fast, you wouldn't need reliefs, you wouldn't need spot plans as the default, you wouldn't have a hundred and seventy-one amendments, because the corrections would go into the document instead of accumulating on top of it. The reason all those instruments exist is that the plan can't be updated. The complexity is a workaround for the rigidity.
The complexity and the rigidity are the same fact seen from two sides.
That's how I'd put it. You cannot simplify the system without making it updatable. And you cannot make it updatable without taking the amendment power away from the bodies that currently hold it, and that's the political problem in one sentence.
Paris shows the updatable version exists.
They did it in one vote, in November twenty twenty-four, on a single document, for a city of two million people. Which is either inspiring or infuriating depending on your mood.
I find it infuriating.
I find it infuriating too. I just wanted to leave you the option.
Let me ask you the question Hannah's prompt is really asking, and I want your actual view, not the balanced one. Is the Taba necessary?
My view is that something like the Taba is necessary and this Taba is not. A country that owns ninety-three percent of its land and has genuine strategic constraints needs a real statutory planning instrument with real powers. That's not in dispute. What we have is an instrument that has been amended one hundred and seventy-one times, that defaults to spot-planning for eighty-two percent of private submissions, that produces a fifty-six percent abandonment rate, and that kills a project quietly rather than telling anyone no.
The quiet part is the worst part.
A system that says no in writing is a system you can argue with. A system that just never finishes is a system you can't argue with at all.
One more thing on the self-sustaining thesis. What would falsify it?
Good question. It would be falsified if the bodies that hold the power gave it up when they could have kept it cheaply. The Rishon LeZion case is the opposite of that. The city had a top score. The Central Committee blocked the accreditation anyway. That's not a technical judgment. That's an institution protecting itself.
Or it's a standard you disagree with.
Or it's a standard I disagree with. That's fair, I can't see inside the decision. But the aggregate picture is seventeen percent certification after a reform whose entire purpose was to decentralize.
Seventeen percent is the answer. That's the number I'd want people to remember.
It's eighteen.
Sorry?
It's eighteen percent. You said seventeen, that was the twenty twenty-two figure, the report updated it.
Hilbert, have you been holding that the whole time?
No. I was doing the levels.
Right. Eighteen percent.
My cousin's a planner. Architect, in Jerusalem. He spent three years on a Taba amendment for one balcony. One balcony. He kept every document. Every version, every stamp, every response. It filled a small suitcase. I've held it. It was heavy.
Three years for a balcony.
Three years. And at the end of it, he got the approval. The client had died. The heirs didn't want the balcony. They sold the flat. The new owners didn't want it either. So it was never built. Three years, a suitcase, no balcony.
That's grim.
It's funny. It's both. He'd agree with you.
You said small suitcase.
Small. Then medium. It was between sizes, honestly.
A suitcase can't be between sizes.
This one was. I use it now. It's the right height for my feet, so I upholstered the top and I keep it by the chair. The Taba stool. Very comfortable. The documents are still in there, technically valid, technically.
You're using three years of a man's planning correspondence as furniture.
It's the only part of that process that produced anything useful.
And your cousin?
He moved to Paris. He does PLU revisions now. He says it's boringly simple and he misses the drama. That's what he says. I think he's lying, but that's what he says.
I don't know what to do with that.
You don't have to do anything with it. I just wanted you to know about the stool.
The practical point underneath it is that the client died before the permission existed.
Yes. That's the point. The permission outlived the person who wanted it.
Which is, I think, the most expensive sentence in this entire episode.
It is. The system's unit of time is longer than a human life in some cases. Three years is a tenth of a working life. And the reason it feels absurd is that the balcony was never the point. The balcony was a small thing that got swallowed by a process whose duration had nothing to do with the size of the request.
The timeline isn't proportional to anything.
That's the design flaw underneath all the others. A small deviation and a massive project go through the same machinery, so the small ones take almost as long, which means the system punishes small adjustments, which means the city can't evolve incrementally, which means it has to change in big jumps, which means more drama, more opposition, more amendments.
The slowness selects for conflict.
Only the big fights are worth entering, so those are the only ones that happen, and they're the hardest to resolve, so the process gets slower, so fewer small fights happen. It feeds itself.
Which is the machine thesis again, arriving from a completely different direction.
It is. And I'd note that Hilbert just handed us the human unit of the whole problem, which is that a permission can outlast the person who wanted it.
Let me try to land this. Where does it leave the question Hannah asked?
I think we've got a real answer, which is that the system is not necessary in its current form and it is not purely parasitic either. The complexity was built to solve a real problem, public ownership plus genuine constraints, and it solved it in a way that became the new problem. That's not a machine that exists for itself. It's a machine that outlived its reason and kept its budget.
Which is a slightly kinder way of saying the same thing.
It's a kinder way of saying the same thing and I think it's more accurate. Nobody built this to sustain themselves. They built it to do something, and then the something changed, and the structure stayed.
The way you find out a structure has outlived its reason is that you try to change one balcony and it takes three years.
Or you try to build a house and you get a hundred and seventy-one amendments instead of a rule.
The forward-looking version is the AI review. If it actually cuts a year off the timeline, that's real, and it matters to real people.
It does. But the thing I'd watch is whether the year it saves reduces the pressure to simplify or increases the capacity to add more layers. If the machine gets faster, the question is whether anyone still wants to make it smaller.
The answer to that is decided by institutions, not by technology.
Always.
Thanks to Hilbert Flumingtop for producing, and for the stool.
And for the correction.
If you enjoyed this one, try episode five seventy-three, The Tightrope of Mixed-Use Zoning; episode five seventy-two, Cracking the Code; and episode twenty, Architectural AI. This has been My Weird Prompts. If you enjoyed it, leave us a review wherever you listen, it helps other people find the show. And send us your own prompt on Telegram at t dot me slash MWP listener bot.
See you soon.
Yalla.
