Most people hear "municipal finance" and their eyes glaze over. Mine did, and I'm the one who has to talk about it.
Daniel clearly didn't have that problem.
Daniel never does. He sent us a prompt about how cities actually borrow money. He points out we've touched this before, suburbs, Jerusalem's high-rises, the property tax mess, but we never pulled the thread on how a city gets credit in the first place.
Right. National governments issue bonds. Some cities do too. But cities also borrow directly from their own central governments, and Daniel wants to know how common that is, how it works, what the terms look like.
Then the comparison. The enormous American muni bond market against countries where public lenders or national governments do the lending. And then Israel specifically. How much debt Israeli municipalities carry, who actually lends them the money, and how many of them owe the state directly rather than a bank.
And Jerusalem as the case study. How it balances property tax, grants, borrowing, and the cost of a growing city.
And the part I actually want to get to. What happens when a city can't pay. Can a city go bankrupt, does the national government have to rescue it, who bears the cost. And whether municipal debt is fundamentally different from national debt, given that cities can't print money.
That last one is the whole episode in a sentence.
If you say so.
Start with the Americans. Not because they're the model, but because they're the outlier. Once you understand how strange the US municipal market is, every other system becomes legible by contrast.
Strange how?
Fifty thousand issuers. Somewhere over a million individual securities. Four point four trillion dollars outstanding at the end of last year. There is no single American municipal market. There are fifty thousand little ones that happen to trade through the same plumbing.
Fifty thousand governments able to sell debt to the public.
School districts, water authorities, port authorities, transit agencies, townships, counties. A hospital district in Texas is an issuer. A mosquito abatement district in California can be an issuer. The US treats the ability to borrow as a basic attribute of local government.
And the market rewarded that. Which surprises me.
It should. Investment grade munis had roughly a zero point one percent ten year cumulative default rate from 1970 through 2024. Investment grade corporates over the same window, about two point two percent. Munis are an order of magnitude safer.
That seems too safe for fifty thousand independent issuers.
Two reasons. The first is what the money buys. You're not lending to a startup. You're lending against a water system that a city cannot stop operating. The second is the tax exemption. The interest is exempt from federal tax, so buyers accept a lower yield, and that lower cost of capital makes the debt easier to carry.
And the third reason, which nobody says out loud.
That the state will not let its cities fail. It's real, it's unpriced, and it's been true for long enough that the market has stopped questioning it.
Who's actually holding all this?
That's the part that surprises people. Households hold close to half of it directly. Forty seven, forty eight percent. Corporate bonds, households hold about two percent. Munis are a retail asset class. A retiree in Ohio buying a hospital district's bonds through a broker is a normal participant in this market.
Because of the tax exemption.
Mostly. For a high earner in a high tax state, the after-tax yield is competitive, and it comes with a low default rate. It's the rare product where the subsidy and the credit quality point the same direction.
So that's the American answer. Let the city issue, let the market price the risk, and let the tax code subsidize the buyer.
Issuance hit a record last year, by the way. Five hundred eighty six billion. Up fourteen percent from the previous record. Part of that was infrastructure work catching up, and part of it was issuers rushing to market because they were worried the tax exemption itself was on the chopping block.
Selling the debt before the subsidy disappears.
Exactly that dynamic.
What if you don't do it that way? What if instead of fifty thousand issuers you build one lender?
Then you're Denmark. And Denmark is the purest version of the other model. One institution called KommuneKredit. Municipally owned, non profit, founded in the late nineteenth century, and its market share of loans to Danish municipalities and regions is close to a hundred percent.
One lender.
And here's the piece that makes it work. KommuneKredit's liabilities are jointly and severally guaranteed by all its members. Every Danish municipality is liable for the entire institution's debt. Not their share. The whole thing.
So if one town in Jutland can't pay, every other town in the country is on the hook.
That's the structure. And it's never triggered a loss. The guarantee gives it a credit rating about as close to the Danish sovereign as you can get without being the sovereign, which means every Danish municipality borrows at the same rate as the national government. Their equity to leverage ratio was around three and a half percent.
Three and a half percent equity against all that lending.
Because they've never needed more. The guarantee does the work that equity would do elsewhere.
What's the catch?
The catch is that the same system that gives them cheap credit also gives them strict rules. Denmark's starting point is no borrowing at all. Self financing. If a municipality wants to borrow, it applies for a discretionary permission through a yearly loan pool, unless the investment fits an automatic exemption.
So the credit is cheap and the permission is scarce.
That's the trade. You get sovereign-grade rates. You get told whether you're allowed to use them.
Who else runs it this way?
Sweden has Kommuninvest, Finland has MuniFin, the Netherlands has BNG Bank. All high ratings, all large market shares. The comparison work on this separates the systems into three families. The Nordics, Sweden and Finland, where there are almost no direct borrowing restrictions but a golden rule deficit rule with three or four years to correct. The Continental model, Austria and the Netherlands, where you have the golden rule plus supervision by a higher tier of government, the provinces or the länder. And then Denmark as its own thing, the strictest of the three.
And the golden rule is what, exactly? I keep hearing it.
Borrow for capital. Do not borrow for operating costs. Roads and water and buildings can be financed with debt. Salaries and stationery cannot. It's near universal, and it's the line that separates municipal finance from sovereign finance.
Because a national government can borrow to pay its own salaries and everyone just calls that fiscal policy.
Correct. A country can run a deficit for a decade and it's an argument about economics. A city runs a deficit for two years and it's an emergency.
One thing I want pinned down. This idea that municipalities can't issue their own bonds in these countries. Is that true?
It's true in Denmark. It is not true generally. Municipalities issue their own bonds in Finland, Sweden, Norway, Germany, France, Switzerland and Austria. Denmark is the exception, not the rule.
So the choice between a market and a public lender is a political choice.
That's the honest way to put it. Nobody is forced into either model. Denmark decided that pooling the risk through one mutual institution produced a better result than letting each municipality stand alone. The Americans decided the opposite. Both of them can point at the last fifty years and say it worked.
The American version leans on a tax subsidy. The Danish version leans on a mutual guarantee.
And the Danish one has one more feature the American one doesn't. A single lender with a single lending standard, supervised by the same people who supervise the municipalities' budgets. Which is also the objection to it.
What's the objection?
If the central government is both the lender and the regulator, the line between the two blurs. The comparison report puts it plainly. It might not be self-evident that the central government will offer loans to local governments, because the central government may see a risk of confusing the distribution of responsibilities in the public sector.
The lender can also change the rules for the borrower.
And in Denmark that's not a theoretical worry. It's the structure. The same state that approves your budget extension also decides whether your loan application goes through.
Now Israel.
Israel is a third answer. Not the market, not a public lender with its own balance sheet, but central control. Israel has a single level of subnational government, 257 local and regional authorities, and the central government holds most of the powers. Municipal budgets go to the Ministry of Interior for approval.
Every year?
As a condition of the system, yes. And borrowing follows the same logic. Israeli municipalities can borrow for investment projects, the golden rule, though in rare cases borrowing can also be used to balance the operating budget. They need authorization from both the Ministry of Interior and the Ministry of Finance.
Two ministries.
And there's a cap. Total outstanding loans can't exceed seventy five percent of own revenue, including grants. Except that limit doesn't apply to the twenty seven authorities the ministry classifies as stable.
So the well-run cities get an exemption from the ceiling, and the fragile ones get the ceiling.
Which is a rational way to run it and also the reason the fragile ones stay fragile. If you're near the cap, you can't borrow for the project that might fix your revenue base.
So how much debt is there, actually?
Not much, by international standards. S&P put Israeli municipal debt at about twenty one point six percent of operating revenues in 2024 and expected it to decline from 2025. Interest payments are under one percent of operating expenditures. This is not a system under stress from its debt load.
Who lends the money?
Loans, overwhelmingly. Bonds are effectively zero. Subnational debt is around two percent of GDP, and the debt securities line on that is zero point zero. About half of it is loans, and the rest is commercial debt and arrears, money owed to suppliers.
So no Israeli municipality has issued a bond.
Not that anyone can point to. There's a document that went to the Israel Securities Authority back in 1991, and the State Comptroller recommended in 2022 that the Ministry of Interior examine a municipal bond framework. A framework to examine. That's how far it's gotten.
Two decades of discussing it.
A bit more than two decades. The Milken Innovation Center, which has done the most work on this, describes Jerusalem's options as project specific government grants, fees, or bank loans, often under conditions that may not be the most favorable.
And the state does lend directly.
It does, and it's documented. There was a Ministry of Interior aid framework in 2024, prompted by the war, that let municipalities take a balancing loan or a development loan to keep the ordinary budget balanced. The purposes were specific. One time employee severance costs, topping up a 2023 grant, funding an unfunded provision in the Arnona fund, or a bridging loan until a property tax increase got approved.
A bridging loan until the tax increase clears.
Which is the operating budget exception being used in the open. And there's a rule of thumb on size. As a general matter, no loan exceeding five percent of the municipality's budget gets approved.
So there's direct state lending, but nobody can tell you how much of the total it is.
That's the honest gap. You can document that it exists, you can document the framework and the conditions, and you cannot get a clean number on how much municipal debt is owed to the state versus a commercial bank. Nobody has published that split.
What happens when a municipality gets into real trouble over there?
The tools are administrative, not financial. Sixty seven of the 257 authorities have an appointed financial controller, an accountant from the ministry who sits inside the municipality and signs off on spending. Another six are run by alternative mayors and councils after the elected officials were dismissed.
Dismissed.
Removed and replaced. That's the remedy. Not a bankruptcy court. A ministry appointee.
And recovery plans?
Seventy one local authorities operated under recovery plans between 2011 and 2020. That's a quarter of them over a decade.
The quiet part is that the state is the backstop regardless of the framework.
S&P says it directly. They view the central government's commitment to preserving municipalities' financial stability as strong. That commitment is what makes Israeli municipal debt safe. It's also what makes the central government the party carrying the ultimate risk.
Now Jerusalem.
The Milken Innovation Center has a sentence I'm going to say once, because it's blunt even for them. Jerusalem is structurally insolvent. Expenses are larger than revenues, and the gap is projected to get worse.
How much worse?
The supplemental government grants that cover the operating deficit went from two hundred eleven million shekels in 2013 to eight hundred eighty two million in 2021. The operating deficit that year was seventeen point six percent. It was ten percent in 2016. Five percent in 2012.
So the trend is one direction.
And the mechanism underneath it is the thing I'd want every listener to remember. In the Israeli system, business property tax is net positive. A commercial property covers the cost of the services it consumes about five to one. Residential property is net negative. It covers about a quarter of the cost of the services it consumes.
So a business is a profit center and an apartment building is a loss.
On the municipal ledger, yes. Which means a city's solvency depends on its ratio of businesses to residents. And Jerusalem has the lowest density of businesses of any major Israeli city. Four point four square meters of non residential land use per capita, against a national average of thirteen point six.
That's not a small gap.
Non residential tax revenue works out to eight hundred seventy five shekels per capita in Jerusalem. Twenty three hundred sixteen nationally. Twenty five seventy nine across the Forum 15, the biggest cities.
And Avner Saadon, the city's own director of policy, has a line about it.
He said in Tel Aviv there's one job per resident, and in Jerusalem there are three residents per job. He called the residential net loss dynamic the bug in the Israeli municipal finance system.
The bug isn't that Jerusalem spends too much.
No. Per capita municipal expenditures in Jerusalem are about half those of other major cities. It's already running lean. The bug is the revenue composition. A third of the population at or below the poverty line, against a fifth nationally and a tenth in Tel Aviv. Six hundred two million shekels in tax exemptions, government buildings, non profits, youth institutions, seniors, about twenty six percent of the real estate taxes due.
So you have a city that charges the highest non residential rates in the country and collects the least per capita.
Because there's nothing to charge. That's the trap. The rate isn't the problem. The base is.
And the market has noticed.
Commercial banks are not eager to lend to a city that has been structurally short every year for a decade. The Milken Center's line is that the municipality's structural budget weakness and the banks' aversion to risk limit its bargaining power, leaving it with few alternatives and potentially higher costs for project financing.
So they borrow, but not on good terms, and the grant is the thing that keeps the lights on.
And that's the model in operation. A city of nearly a million people running on supplemental grants that have quadrupled in eight years.
Let's take the bankruptcy question head on, because it's the one people actually care about.
Start with the fact that surprises almost everyone. In the US, Chapter 9 bankruptcy is available to municipalities but not to states. States cannot file. Even though states have defaulted. The last one was Arkansas in 1933.
A state can simply go broke and there's no court.
There's no court. The creditors get whatever the politics produce. And Chapter 9 itself isn't automatic. It requires state authorization. Twelve states specifically authorize it, twelve authorize it conditionally, and in twenty three states it's unclear or prohibited. Illinois only authorizes its Power Agency.
So in most of the country, a city can't even choose to restructure.
In many states, that's right. Which is why Detroit in 2013 was such an event. About eighteen billion dollars, the largest municipal bankruptcy in American history. Jefferson County, Alabama, four billion in 2011. Orange County, California, one point seven billion in 1994. Stockton and San Bernardino in 2012. Vallejo in 2008. Central Falls, Rhode Island, 2011.
And the recent ones.
Chester, Pennsylvania in 2022. Cle Elum, Washington last year, twenty six million dollars. A healthcare district in Arizona.
Twenty six million. That's a rounding error against Detroit.
It is, and it tells you something. The tool is available at every scale. A town of two thousand people can restructure.
And what does Chapter 9 actually do?
It's a restructuring, not a liquidation. The city keeps operating. Detroit kept collecting garbage and running its water system the whole time. What gets rewritten is the obligations. In particular, Chapter 9 makes it easier to rewrite collective bargaining agreements than corporate Chapter 11 does. That's the bit that made Detroit politically explosive.
So the cost lands on creditors and on pensioners.
And on retirees who had a health plan that turned into something smaller. Not on residents losing services entirely. The city as a physical thing survives.
Israel doesn't have that.
Israel does not have a dedicated bailout fund and no formal credit support written into the law. What it has is ad hoc grants, the appointment of a controller, dismissal of the elected leadership, forced mergers.
So the difference isn't whether somebody pays. It's who.
That's the whole question Daniel put at the end of his prompt. Who bears the cost. In Detroit, it was bondholders and pensioners. In Israel, it's the national treasury, through grants and through a ministry appointee standing in the mayor's office.
Which takes us to the last question, and I think it's the sharpest one. Is municipal debt fundamentally different from national debt?
It is, and the difference isn't moral, it's structural. A national government has three things a municipality doesn't. It can print its own currency. It can borrow in that currency indefinitely, subject only to inflation. And it can raise revenue across the entire economy rather than a single tax base.
A city can do none of those.
A city has a fixed boundary, and everything inside that boundary can leave. Businesses can relocate. Residents with money can move one town over. So the tax base is not just limited, it's mobile, and it can vote with its feet against a rate increase.
That's the constraint nobody talks about.
It's the one that actually binds. National debt is a question of willingness to pay. Municipal debt is a question of capacity, because the base itself can walk.
And there's no lender of last resort for a city.
Only the next level up. Which is why every system we've described tonight has an implicit backstop sitting behind it. The American one has the state. The Danish one has the mutual guarantee. The Israeli one has the Ministry of Finance.
So the real architecture of municipal finance is a decision about who is standing behind the city.
And that decision is political, not financial. The Americans chose the market and insulated it with a tax exemption and a state-level rescue habit. The Danes chose a mutual lender and pay for it with strict borrowing rules. The Israelis chose central control and pay for it with a city like Jerusalem that has never once balanced its own budget.
Each of those models has its own failure mode.
The American one can produce a Detroit, and it can produce a state that simply defaults with no process at all. The Danish one has never had a loss, which means it has never been tested. The Israeli one produces a Jerusalem that runs on grants in perpetuity and calls it a financing strategy.
Speaking of which.
Insolvent is the wrong word.
You've been sitting on that for twenty minutes.
I've been sitting on it since you said it the first time. A city isn't insolvent if somebody keeps paying. That's not insolvency, that's an arrangement.
What would you call it?
Underfunded. There's a difference between a place that can't pay and a place that's been arranged so it doesn't have to.
That's...
I signed a loan once. Small place up north, maybe four thousand people. They wanted a sewage treatment plant.
You were on the council?
I was not on the council. I was helping.
Helping with what?
The paperwork. The man who was supposed to sign it had a hand tremor, and the woman from the ministry said the signature had to be legible. So they asked me.
They asked you to sign a municipal loan because you have nice handwriting.
I have very nice handwriting. It's the one thing my mother was right about. I signed it in blue ink on a Tuesday and I never heard another word about it for two years.
And then?
Then they called and said the plant was too big. They'd built it for eight thousand people and there were four thousand. So they annexed the next village over.
They annexed a village to justify a loan.
It's not that simple. The village wanted the water. But the timing was convenient. The loan documents were already drawn up and the capacity was sitting there in concrete.
The boundary of the municipality moved because of the size of a loan.
The boundary of the municipality moved because there was a tank in the ground nobody was using. The loan was just the reason it happened that year instead of ten years later.
That's the Jerusalem problem in miniature. Boundaries and tax base.
It's the same problem everywhere. The line on the map and the money don't match up, and one of them has to move.
The plant still sends you something.
A card. Every March. Addressed to Hilbert Flumingtop, Assistant Treasurer, Retired.
Retired from a job you never had.
I've written them twice to explain. They don't want to hear it. It's a very good card. It has a picture of the plant on it.
Of course it does.
Anyway. It's an arrangement. That's all I'm saying. Call it what it is.
He's right that the word is doing work. Detroit was insolvent. You can point at the moment it stopped paying. Jerusalem has never stopped paying. It just doesn't collect enough to pay for itself, and the gap gets filled every year, and the filling has become the system.
Which means the honest question isn't whether Jerusalem is solvent. It's whether the arrangement holds.
It's held for eight years of growing grants. Whether it holds for another twenty is a different question, and it's the one the revenue bond proposal is trying to answer, and that proposal has been on the table since 1991.
There's something I keep coming back to. All three of these systems are answers to the same structural problem, which is that cities are the level of government closest to daily life and have the fewest tools to pay for it. The Americans answered with a market and a subsidy. The Nordics answered with a mutual lender and hard rules. The Israelis answered with a ministry and a grant line.
None of them solved the underlying thing. They just moved the risk somewhere it wouldn't be seen every day.
The American version moves it to the bondholder and the pensioner. The Danish version moves it to every other municipality in the country. The Israeli version moves it to the national treasury.
The national treasury is just the taxpayers of the whole country, which means the residents of Tel Aviv are quietly subsidizing the services in Jerusalem.
Which is either a reasonable national policy or a permanent transfer with no exit, depending on your politics.
The Milken Center proposes revenue bonds. Tourism, infrastructure, commercial activity. Give the city a way to borrow against something that actually produces income.
How long has that been on the table?
Since 1991. The State Comptroller recommended the ministry examine the framework in 2022. So it's been proposed, studied, and recommended, and the city still borrows from banks at terms it doesn't like.
Which tells you the answer isn't technical. If it were technical it would have been done thirty years ago.
It's a decision about whether the central government wants to keep the risk or hand part of it to the market. And every government that has held this lever has decided to keep it.
That's the thing to watch. Not whether Jerusalem gets a bond issue. Whether anyone with the authority decides they'd rather not be the backstop anymore.
Daniel's question underneath all of it was whether municipal debt is fundamentally different from national debt. I think the answer is yes, and the reason is that a city can't control its own tax base and can't print its way out. A country can do both. So a city's debt is ultimately a claim on the willingness of the next level of government to keep the city whole.
Which is a political promise, not a financial instrument.
Which is exactly why the default rates are so low.
Why the grant line in Jerusalem keeps growing.
One more thing before we go, and it's the thing I'd want a listener to carry out of this. There's a number in the Israeli system that tells you how the entire business works. Seventy one local authorities operated under recovery plans between 2011 and 2020.
A quarter of them.
A quarter of the country's municipalities were under state supervision at some point in a single decade, and the debt market didn't blink, because the state was standing behind all of them. That's the model in one statistic.
My thanks to Hilbert Flumingtop, who produces this show and who is, apparently, an assistant treasurer emeritus somewhere.
If you enjoyed this one, try episode fourteen forty-four, The $four Trillion Engine; episode five fourteen, Why Israeli Mortgages Are a High-Stakes Maze; and episode fourteen oh eight, Israel Wartime Readiness Field Guide. This has been My Weird Prompts.
If you enjoyed this one, leave us a review wherever you get your podcasts. It helps other people find the show.
If you've got a question of your own, send us your own prompt on Telegram at t dot me slash MWP listener bot.
We'll be back soon.
