Which acre pays for the pipes?

This past September, in an effort for my friend group to get to know one another better, we held a PowerPoint party. Each of us picked a topic; most chose from our educational background or some general special interest. I couldn't help but harken back to one of my favorite subjects, my minor and my policy obsession: housing and urban design.

Here in Chicago there is a small contingent of folks who would love to see Lake Shore Drive torn down and replaced with passenger rail. I admit, I think it is a sleek and futuristic vision for the city, and the lakefront would be better served by rail than by a six-lane freeway. Do I think it's a viable project? No. The math doesn't add up. The lakefront doesn't need added service, and the money would be far better spent connecting the areas rail doesn't reach and closing Chicago's spoke-and-wheel into a nodal system. That argument is its own post, and it's coming.

If we loaded a map of Chicago into SimCity, I would absolutely tear down Lake Shore Drive and lay rail, because I am notoriously good at making money in those games and could afford a vanity project. Why? Because I follow simple planning rules that Americans have been failing to abide by for nearly four generations. The first rule is a ratio: the land serviced by tax dollars against the privately owned land that pays the taxes to maintain it. A mile of road and pipes costs about the same wherever you put it. What varies is the size of the base it serves.

Let me show you what I mean. Two commercial properties, same county, about two miles apart.

The first is a block of downtown Oak Park: Lake Street to the north, North Boulevard to the south, Marion and Forest on the sides. Nineteen parcels, three and four stories, shops on the ground floor with offices and apartments above, most of it standing since before 1930. It is the kind of block every suburb in Chicagoland wishes it had. The whole thing covers 3.4 acres.

The second is the Walmart Supercenter at 7500 Roosevelt Road in Forest Park. One store, one parking lot, 17.7 acres. Five of those Oak Park blocks would fit inside it.

Now ask the only question a local budget actually cares about: which one pays for the pipes?

The Oak Park block was billed about $1.35 million in property tax for 2025. The Walmart was billed about $753,000. Fine, the block pays more; it has more stuff on it. Divide each by the land it sits on and the gap stops being polite: roughly $394,000 per acre against roughly $43,000. Nine to one. Same county, same state equalization factor, two villages with similar tax rates. Oak Park's composite rate is a bit higher, 12.2 percent against Forest Park's 10.8, and that explains about one of the nine. The other eight are the land.

I work with local campaigns and county officials, and this is the argument I wish more of them would make, because there is nothing ideological in it. It is levy math.

I am not the first to notice this, and I would be a fraud to pretend otherwise. Joseph Minicozzi's firm Urban3 has run this comparison in hundreds of American cities since 2012, starting with a renovated downtown building in Asheville that produced about $250,000 per acre in county property tax against $8,000 for the mall out on the highway. Smart Growth America reviewed seventeen local fiscal studies and boiled them down to one sentence: compact development generates ten times more tax revenue per acre, costs 38 percent less in upfront infrastructure, and costs 10 percent less to serve with police, fire and ambulance.

So why doesn't your county measure it this way? Because it compares projects by total value, and by total value the big box looks fantastic: one large assessment, one ribbon cutting, one happy headline. But infrastructure is not priced per parcel. Roads, water mains, storm sewers and signals cost by the mile, and a mile of pipe runs past every acre whether that acre is producing $394,000 or $43,000. Per-acre yield is the number that tells you whether the land along a mile of pipe can pay for the pipe.

The pushback I get, and it is a fair one: dense blocks carry heavier infrastructure. Bigger mains, older systems, more to fix. Doesn't that eat the difference? It narrows it. It comes nowhere close to closing it. Halifax, Nova Scotia ran the cleanest version of this study I have found, costing out nine municipal services across every settlement pattern in the region. Annual cost per household came to about $1,400 in the dense urban neighborhoods, about $3,500 in standard suburbs, and over $5,200 in rural subdivisions, in 2003-04 Canadian dollars. The dense neighborhoods carried the heavier pipes and still cost less than half as much per household to serve.

Costs scale by the mile. Revenue scales by the acre. Density puts more tax base on every mile you are already obligated to maintain. That is the whole argument, and it fit on one slide.

What a board can do with this

1. Ask your assessor for tax yield per acre, not per parcel. The data is public; mine came off the county's open data portal. One afternoon shows you which land is carrying your county and which land is being carried.
2. Treat your pre-1940 downtown blocks as revenue infrastructure. They are usually the highest-yield acres you govern, and in most places the zoning code makes it illegal to build another one.
3. Price the maintenance tail before approving the next arterial strip. The ribbon cutting books revenue this year. The road, storm sewer, and signal maintenance land on budgets ten to thirty years out.
4. Notice that farmland preservation and fiscal discipline are the same argument. Every acre of low-yield development consumes farmland and adds a maintenance liability. The fiscally conservative position and the conservation position point the same direction, which is why this wins in red counties.

County and village boards control land use in most of America outside the big cities. The boards that learn to read their budgets per acre stop approving their own future deficits.

The joke is that one man gets to plan a whole town for himself. Chicago's version of the joke is that we keep handing one developer sixty acres and calling it a neighborhood. It's time for our cities to stop planning for Billionaire Weenie Town with massive megaparcels only accessible to large capital. Cities ought to be providing more small, traditional, granular lots that are accessible to smaller-scale developers and families.

Jonathan Konkol perfectly encapsulates this point in his 2020 piece for Plan Design Xplore, Why Not Incrementalism?:

"This way of building is, however, an anomaly in the history of human habitation. While Popes and Caesars had the power to reshape districts, almost all building in cities has been incremental and small scale in nature. Until the rise of corporate investment banking, only the church and the crown had the power to build this way... Most of our cities were developed on a grid system made up of modules of 50x100 foot lots. It's far easier for a small local developer to purchase one or two of these and put up a modest building... When lots of parcels are consolidated, only the big national builders, fueled by institutional capital, can compete."

Chicago has been running Jonathan's experiment in reverse for fifty-five years, and I can show you the scoreboard.

The 62 acres at Roosevelt and Clark have been vacant since Grand Central Station came down in 1971. South Works, 415 acres of lakefront, has sat empty since the mill closed in 1992. Block 37, in the middle of the Loop, sat cleared for twenty years. The Finkl steel site on the North Branch has been empty since 2012. Every one of them was assembled into a single parcel, entitled as a single planned development, tied to a single developer's financing, and handed its own public subsidy. And every time that one developer's balance sheet or thesis broke (a corruption case, a steel company walking away, a REIT collapse, an office market, a pandemic) the entire site stopped, because nobody else had standing to build on any part of it. The shocks were different every time. The structure was the constant.

Lincoln Yards is the one you can watch in real time. Sterling Bay bought the Finkl site in 2016 and won approval in April 2019 for a $6 billion, 14.5 million square foot plan with up to $1.3 billion in tax increment financing behind it. It completed exactly one building, a life sciences tower that sat empty. The lender took the northern half in 2025. The site is now split three ways by foreclosure sale, and on September 23 the City Council approved $201.6 million in public money for the roads, parks and bridge: roughly a fifth of the subsidy the original deal carried, seven years late.

Meanwhile, next door, the ordinary blocks of Lincoln Park and Bucktown did what ordinary blocks do. Between 2016 and this month they permitted about 77 new buildings on individual lots, roughly $40 million of private construction, with at least one new home permitted every single year, 2020 included. The 53 acres beside them permitted one building. Same pandemic, same interest rates, same city. The shocks that killed the megaproject never reached the neighborhood.

This is the whole difference between one balance sheet and many. Plat the northern 31 acres of Lincoln Yards on Chicago's standard 25 by 125 foot lot and you get about 280 lots. At the surrounding blocks' mix that is roughly 500 homes; at three-flat density, more than 800; assemble the corners and the arterials for mid-rise, the way every pre-war Chicago neighborhood actually grew, and you are at two thousand or more. Foundry Park's approved plan is up to 3,737 units on the same land, and I am not going to pretend otherwise: a plat trades peak density for speed, resilience and local ownership. What it never does is sit empty. The plat starts producing homes in year one and never stops. The megaproject produced zero homes in fourteen years and is now scheduled to finish in 2036, twenty-four years after Finkl closed. We need the money to start rolling now, not in ten years.

Chicago already knows how to do this, as long as the land is cheap. The Large Lots program sold vacant city lots to the neighbors for a dollar. The Missing Middle initiative sells clusters of city lots to small developers for a dollar apiece, with construction subsidy, for two-flats and three-flats; its September round drew 34 applications for 30 lots in West Englewood and South Chicago, and the program had already put more than 300 homes across 106 buildings in motion. So the city sells small lots to small builders on the South Side and hands 53 acres to one developer on the North Side. Why does that knowledge stop at the river?

This post is arguing that our government ought to work toward providing and favoring small developers with more land opportunities, and to recognize that granular platting is the more reliable path to the long-term development, and the eventual redevelopment, of our land.

One more thing the permits show, and it is the reason for the next post. On Lincoln Park's own blocks, incrementalism mostly meant two-flats torn down for single-family houses. Small lots solve the problem of who gets to build. They do not decide what gets built. That is a decision a neighborhood has to make on purpose, and right now nobody asks it to.

Credit (interim wording): The method here follows Jonathan Konkol's 2020 study of Portland's Lloyd Cinemas site; the Chicago application and the numbers are mine.

The median home in America now costs five times the median household income. In 1985 it was three and a half times; through the 1990s, about three. What does that mean? Americans starting from scratch, going from renter to homeowner without family help, have to save and dedicate larger portions of their income to it than any generation before them.

Whether my fellow left-leaning Americans like to acknowledge the reality of market dynamics or not, we have a supply issue. There are simply more people than there are houses to put them in, and while the economy runs hot, a limited and inelastic housing supply will keep rising in price until we let the market adjust to demand. Some Americans think we can deport our way to lower housing costs. For the rest of us, it's time to make adjustments: for neighborhoods to pull up their big-boy pants, open up the zoning code, and make clear, easy paths for new higher-density housing in all neighborhoods, without the lengthy design review processes that do real harm. The current model is hurting young Americans and our growth as a nation.

We still live in a democratic system, and asking neighborhoods to give up their power is not something that will happen easily or overnight. We can take action like the State of Oregon and ban single-family zoning in every city, allowing fourplexes on any lot, or we can look for more granular approaches to adding higher-density housing in older neighborhoods while preserving the high-quality historic building stock we already have.

After all, we aren't building any new Victorians. It is a shame to see a 130-year-old Victorian mansion torn down and replaced by a fourplex while a 1970s snout-nose, garage-first single-family house stands untouched next door. Decades of NIMBYism have proven that we can't trust neighborhoods to respond responsibly to higher demand for housing; many neighbors work deliberately to hold up new projects with the goal of limiting supply in high-demand areas, and so increasing their own property wealth. So now state legislatures are stepping in to make the supply changes themselves, at larger and less surgical scales than neighbors and cities could have managed on their own, and what we are left with is unhappy neighbors and uglier outcomes. What states ought to demand instead is that every neighborhood allocate new housing within its plans in line with past and predicted population growth, rather than pushing growth to the edges of cities, eating farmland and adding traffic.

Neighborhoods ought to be forced to accept change, growth and new housing, no matter how quaint or "historic." But neighbors ought to be armed with the tools to choose how and where that growth goes, in ways that preserve what their communities value most. We need a democratic system that lets neighbors plan the future growth of their community, lock it in, and then step out of the way so that developers have a clear and seamless path to delivering the housing we need.

Here is the distinction that makes that possible, and it took me an embarrassingly long time to see it. Neighborhoods cannot be trusted with whether; the share has to be assigned from above. They can absolutely be trusted with where and what, once the number is fixed. One Saturday in Portland proved it.

Dynamic Density was developed by Rick Potestio and Jonathan Konkol of ReUrbanist Collaborative. The three of us found each other in Portland by accident rather than design, and I argued with them through every version of it; the tool is theirs. [Interim wording.] It works in three steps. The neighborhood inventories its blocks and decides what is off-limits: buildings, trees, businesses, whatever it would actually fight to keep. It receives a specified number of new homes, its share of anticipated growth. And it places them on the remaining properties by choosing from a menu of pre-war multifamily building types that Rick and Jonathan measured off real buildings, walking Portland with cameras and tape measures. It is a board game with a budget.

In February 2024 a couple dozen neighbors in Portland's Irvington Historic District played it with Legos. Three teams, 600 units each, on a map already color-coded for contributing and non-contributing buildings. One team got mostly duplexes and fourplexes, one got mid-sized buildings, one got a few towers. Every team sacrificed the non-contributing buildings first; nobody had to tell them to. The small-building team had the hardest afternoon and the most demolitions, and concluded their scenario would not pencil in a neighborhood that expensive. The teams with a few taller buildings touched fewer sites and saved more. Jonathan's summary is the hinge of this whole series:

"One can minimize changes in the scale of buildings, their bulk and height, or one can minimize demolitions of existing buildings by keeping the footprint of change small and going up. Mathematically, one cannot do both."

That is my Victorian and my snout-nose house, decided on purpose instead of by accident. It is also the answer to the Lincoln Park teardowns in the last post: left to lot-by-lot economics, "gentle" infill eats the two-flats. Allocation has to be by type, at neighborhood scale, or the market allocates for you.

Chicago already lets neighborhoods decide, by the way. It just does it through one person, after the fact, one parcel at a time. Aldermanic prerogative is the unwritten custom that gives the local alderperson final say on zoning in their ward. Half a mile from Lincoln Yards, Sterling Bay's 1840 N. Marcey proposal took a year and a public fight before a negotiated deal passed in July 2025, with no reference to any neighborhood-wide target, because none exists. The same month, a citywide ordinance allowing coach houses and basement units stalled; it passed in September only after being converted back into ward-by-ward alderperson approval with a menu of optional restrictions. That is neighborhood control as Chicago practices it: the right authority, at the wrong end of the process, bound to no number. Dynamic Density moves that same authority to the front, in public, against a target. The alderperson becomes the convener of the allocation instead of its gatekeeper, which, if I were an alderperson, is the job I would rather have.

Irvington's afternoon worked because the map already existed. Chicago's does too. The Chicago Historic Resources Survey spent thirteen years rating roughly 500,000 pre-1940 buildings by color, and the red and orange layer sits on the city's open data portal right now. Every landmark district designation report lists contributing and non-contributing properties parcel by parcel. The inventory step is done citywide; we just never did anything with it. Take Logan Square's boulevards: 265 acres of graystone two- and three-flats built between 1890 and 1930, a Chicago Landmark since 2005, the Blue Line running straight through it, and a neighborhood association that already argues about exactly these questions. The workshop it would need has its map, its building types and its blocks. What it lacks is the number.

Nobody has told Logan Square its number. The next post is about who does, and what happens the day after.

Life as a housing developer is a never-ending word problem. The inputs and outputs in a proforma change daily, and years can be spent in design review, court battles and public hearings before ground is broken. All of that drives up the cost of new development, and the cost is passed on to the people who will eventually rent or buy.

What if neighborhoods front-loaded the design review process for developers? In the last post we looked at Dynamic Density, a model for letting a neighborhood allocate its share of new housing and update its own zoning. It forces cities to plan for growth, and it gives neighbors the power to shape it. Once that density is allocated, the neighborhood can go one step further and choose a catalog of building types it has already approved; the village then commissions stamped plan sets for each, so a developer who builds to the catalog gets a fast track through permitting. Design, review and approval happen once, in public, before anyone buys a lot. By the time a developer shows up, the where, the what and the number are already decided. The developer's job becomes accepting a decided project. Call it the pickup.

None of the parts are new, and I want to be exact about that, because the whole point is that none of them have to be. Neighborhood allocation is Rick and Jonathan's. Pre-approved plan catalogs exist: South Bend, Indiana launched one in 2021 and now offers seven plan sets from an accessory unit to a six-plex, with 18 homes built by last spring. Public-plans-and-services-first exists: Copenhagen's development corporation lays the streets and the Metro, then sells serviced plots. And public land on long ground leases exists. Jonathan's 2019 proposal for Portland's golf courses put every new building on a 99-year ground lease on land the city kept, and New York has run that model for fifty years at Battery Park City, where a state corporation owns 92 acres on ground leases, has sent $461 million to affordable housing since 2010, and committed $500 million more in 2024. A ground lease is a lease. The public body that leases keeps the appreciation its own pipes created, which is the Part one argument with a deed attached.

What has not been done is welding those parts together with one more piece, the per-acre number from the first post, as the approval criterion. The board approves a site because the types the neighborhood chose will clear a stated tax yield per acre and the maintenance tail is priced, and it recovers what it spends on streets and pipes through land it keeps rather than through one developer's increment. That weld is the whole claim of this series.

If that sounds like a lot to ask of a village, look at what Chicago already asks of itself. The 62 acres at Roosevelt and Clark were approved in 2018 for a $7 billion plan. In 2019 the city promised $551 million in future tax increment to reimburse the developer for a Red Line station, Metra tracks, streets and a seawall. Nothing got built for six years. The site moved in 2025 only when a private owner agreed to fund a soccer stadium outright, and this July the city rewrote the 2019 deal: up to $425 million in tax increment, now aimed at streets, track work, a river wall and a 1,200-space garage the city will own. Public infrastructure first, private construction second, on a plan already decided. That is the pickup in everything but the land. The difference is that the developer keeps the 62 acres and the increment, and the public gets a parking garage. Foundry Park, on the Lincoln Yards site, is the same movie: $201.6 million of public money for streets, parks and a bridge, seven years after the first developer promised to build them.

None of this needs new state law, which surprised me. An Illinois village already zones and plats its own land. Tax increment financing, the tool that funded Lincoln Yards, The 78 and Foundry Park, runs 23 years and pays for exactly the things the pickup needs: land, streets, pipes, site preparation. A Special Service Area lets a neighborhood tax itself for its own improvements or bond against them, which is the Illinois version of the neighborhood development corporation Rick and Jonathan proposed in 2019; Oak Park already has one on its tax bill. The Cook County Land Bank can hold title. The per-acre floor is a board resolution. A village board could adopt the whole thing on a Tuesday night. One correction the earlier posts owe: in Illinois, county boards zone only unincorporated land. Every village downtown is zoned by its own trustees, and they are the audience I am writing for.

Three seams, named, because a trustee will find them anyway and I would rather be the one pointing. First, the arithmetic: almost no town computes tax yield per acre, and the model needs that number before it spends a dollar. Second, timing: a TIF or a bond against future lease income is a real risk if lots are slow to go. Third, financing: a serviced lot with a stamped plan strips the soft costs off a small builder's proforma, but it does not lend to them. The honest version is that the public takes the timing risk in exchange for keeping the land, and that it already takes that risk today, later and worse, without keeping anything.

So here is the pilot. One village or small city names one site it already owns or can hold through the land bank: stalled under at least one prior master plan, next to pre-1940 fabric the Historic Resources Survey has already mapped, somewhere between five and forty acres. The council names the site's share. The neighborhood association runs the allocation on the survey map with five building types measured from the town's own streets. The village plats it on standard lots, prices the pipes, and adopts a per-acre floor by resolution. A TIF or a Special Service Area funds the streets. The village or the land bank holds title. Lots go out on ground leases with stamped plans, two at a time, to whoever builds the approved type first, with the incentive aimed at early adopters and local builders. The number the board approves is dollars per acre, not units.

I will run the workshop and the per-acre arithmetic for the first village that wants to try it.

Four slides, four posts, one afternoon with friends that turned into a plan a village could adopt by resolution. That is what the party was for.