Skip to content

Dream Houses

Keep building, just a bit more!


Thomas More’s Utopia has become the archetypical presentation of earthly paradise, famously detailing a prospective society in which reason and progress rule all. Despite protestations of realism, the best way to read Alex Hemingway’s essay “To Solve the Housing Crisis, We Have to Increase the Housing Supply” is as a particularly insipid installment in this utopian tradition. Published in Jacobin in September, it is furthermore an exercise in imaginary promises undertaken by a latter-day adherent of the capitalist injunction to pursue growth at all costs. That this piece appears in Jacobin, an ostensibly “socialist” publication, goes a long way towards burying its call for capital accumulation under the thinnest veneer of left politics at its most bleeding-heart chauvinist.

Hemingway begins by declaring a supposed fact: “The truth is there is no way out of the housing crisis without a tremendous effort to end the chronic shortage of homes, which helps drive up rents.” He then goes on to list five reasons why “the issue of supply should matter even—maybe especially—to those of us who focus most of our efforts on expanding public and nonmarket housing”: (1) a low vacancy rate benefits landlords; (2) large investors “thrive on the housing shortage”; (3) the previous two points are proven true by academic studies; (4) which means that building must continue, without end, until rents drop; and (5) any attempt to secure rights for renters against landlords must come after the housing shortage is ended in accordance with reason number four. I can sum these five points up with two: (1) capitalism can alone solve its own crises; and (2) it should be allowed to do so, regardless of how long it takes, before we even consider intervening in other ways.

The magnitude of this shortage of homes (or, really, a supply deficit) is not identified until later in the essay, when he cites a Canadian Mortgage and Housing Corporation (CMHC) report that states the province of British Columbia needs to add five hundred seventy thousand additional residential units above current projections between 2021 and the beginning of 2030 to return to the level of affordability (measured as rent or monthly mortgage payments as a share of total household income) last seen in 2003–2004[*]. The timing of this is important, as it is just before Canada began experiencing its own, slow-motion version of the U.S. subprime mortgage crisis, reflected in the dramatic and sustained rise of residential property prices—likewise spurred on by low interest rates and mortgage lending. At the same time, rent prices have been on a tear, increasing well ahead of wages. In effect, the CMHC proposes turning back the clock nearly two decades in an attempt to shift the blame, to mutate a crisis of rapacious financial lending into a crisis of construction—a bit like fussing about a splinter while leaving a gaping chest wound untreated.

The theory of explosive growth as a rising tide that lifts all boats simply cannot be made to fit real conditions, regardless what form it takes.

Let’s work out the numbers here. Suppose a province-scale residential construction project of the type Hemingway espouses is suddenly implemented. If we take the number of housing units completed in 2019 as a suitable projection, the CMHC is calling for around sixty-three thousand on top of that, or just over one hundred thousand new units per year, every year, until 2030. Disregarding the immense amount of construction this would require in new utilities, roads, etc., and focusing on residential construction alone, fulfilling this called-for expansion of units would require an increase in labor productivity (which is the ratio of real value added to total hours worked) from 2019’s 53.10 CAD per hour to 138.5 CAD per hour or, assuming labor productivity remains the same, 499 million labor hours a year, which, given a forty-hour workweek, would require over two hundred sixty thousand residential construction workers. The wage bill, assuming an annual salary of 63,375 CAD, would be 17 billion CAD. This would also contribute to considerable cost price increases in construction inputs (like lumber, concrete, etc.) due to the increase in demand for such materials, though the plan calls for no increase in manufacturing. In light of this, Hemingway’s claim that this “housing for all” scheme is “achievable” and “affordable” now seems to be a bit less than pragmatic.

All this aside, Hemingway’s proof that new construction lowers rents depends on a theory that the housing market behaves on the principle of “filtering.” This model of housing mobility assumes that all households (or just people) in a given market respond to new supply relatively instantly, and when a new luxury unit is produced, the residents of what was formerly the best unit move from their current unit to this new, highest-quality unit (quality being measured here in “housing services,” whatever that means). This then prompts the residents of what was formerly the second highest-quality unit to move into the formerly highest-quality unit, and the great “moving chain” continues, from the top on down, until everyone has found themselves a slightly nicer place to call home. This hermit-crab game assumes that all players know the quality of their units, that they can all move instantaneously, and that there is no segregation or selectivity in the leasing process (and thus, more or less, no leases and no landlords). It also assumes that income, and thereby the amount of rent paid, is not a factor in a household’s choice to move into a better place. It’s all about maximizing “housing services,” baby.

Hemingway exclaims that the filtering principle is backed up by “rigorous academic research,” citing the working papers “City-wide Effects of New Housing Supply: Evidence from Moving Chains” by Cristina Bratu, Oskari Harjunen, and Tuukka Saarimaa and “The Effect of New Market-Rate Housing Construction on the Low-Income Housing Market” by Evan Mast (neither of which happen to be to be peer-reviewed). All these working papers have to do to prove Hemingway’s argument is posit that an increase in housing supply of any type—especially luxury residential—has rent reduction effects within a particular market.

Simply put, they can’t prove it. For example, while Bratu, Harjunen, and Saarimaa’s working paper states, rather stunningly, that for one hundred new luxury market-rate units, twenty-nine additional units get created through vacancy in the 20 percent lowest income zip codes and sixty units in the lowest 50 percent. The language they use is important: though Hemingway claims their research shows new building increases affordability, these researchers he cites are not claiming this at all. What they do claim is that more existing units of a worse quality and thus a more accessible price for lower-income households will become available if only highest-quality (read: luxury) units continue entering the market. This leaves the researchers with the useless task of proving the existence of an imaginary “moving chain.”

They are almost able to do this by undertaking an extensive abstraction of the data, but not quite. For example, when confronted with the issue of a lack of origin unit data on new arrivals into a residential neighborhood (that is, there’s no way to know where exactly someone moved from, and thus prove that their initial statements are correct) the researchers instead claim that any moves from a particular geographic tract with a lower median income at all within a particular time window counts as proof of the “moving chain” in action; put more simply, having only an occupied luxury unit as given, the researchers then look citywide to find any incoming moves into the next-worst housing category, and the next-worst after that, and so on down the line. This should be very easy considering each neighborhood assigned to a particular category contains many thousands of units to choose from, if not more. But even with these extremely forgiving conditions, Bratu, Harjunen, and Saarimaa’s statistical witchcraft shows that nearly 90 percent of chains end before reaching the lowest-income households. So: even with the broadest possible definition of proof required to fulfill their claim, they still can’t show that new luxury construction actually enables people to move into better apartments. What’s more, they have no way of actually tracking reductions in rent because neither household rent payments nor incomes, the ratio of which is the commonly accepted definition of affordability, are tracked. Somehow, even with this lack of data in support of a useless thesis, we are meant to believe this constitutes rigorous statistical proof that rents will tend to become affordable if new buildings are constructed.

A look at the real world shows a rather different picture, as it often does when it comes to economic theory. Travis County, Texas (part of the Austin–Round Rock–Georgetown Metropolitan Statistical Area), is one of the fastest-growing counties in the country in terms of new home construction as of July 2021, having constructed a whopping 25,693 units in the last year. So we may expect, if Hemingway et al. are correct, the rents would have fallen correspondingly, right? Of course, they did not—in fact, asking rents in Austin increased 40 percent over that same period, and have continued to rise well into 2022—even though the vacancy rate is a “healthy” 8 percent. What gives? There is an obvious answer here: the theory of explosive growth as a rising tide that lifts all boats (or, trickle-down economics) simply cannot be made to fit real conditions, regardless of what form it takes, be it doggedly idiotic calls to build at all costs, be it more subtle academic “proofs,” or otherwise; all are constantly shamed when exposed to reality, where landlords enjoy unilateral power to set rent prices on market-rate units, where new construction usually rises from the rubble of minority neighborhoods, and where the benefits of new construction do not accrue to all, but almost entirely to the owners.

Hemingway, to his extraordinarily meager credit, does also call for the production of public housing—which partially explains why this drivel might have ended up in Jacobin. But here, his proposals are just as myopic as elsewhere. Let’s take up another hypothetical: “housing for all!” has gripped the masses, resulting in the creation of a public construction corporation (which the author advocates for elsewhere) at the national level, bequeathed with the explicit mandate to build housing constantly, irrespective of fluctuations in input costs, land acquisition prices, or labor. Basically, then, it is to function as an “angel” building corporation, which, unlike all others, never needs to make a profit—ever—thanks to the magic of infinite public money. Let’s say that this firm constructs one thousand new units a month, to be rented by the corporation directly at rents corresponding to a certain affordable standard. Existing owners (be they real estate investment trusts with large portfolios, smallholder landlords, or otherwise) would experience this new public competitor as a lethal foe, with each new unit constituting the loss of a potential consumer of one of their own units.

If only this were what Hemingway had proposed! Curiously, after delivering his triumphant call for massive public construction, he turns immediately into a meekly pragmatic clerk, hat in hand, wholly unconcerned with the fundamental question of the distribution. His only concern is the construction of units, which, if done by a public authority would equal the extraction of public money both from capitalist profits and household revenues (wages) and its transformation into capital housing assets, thereby constituting a form of theft in which either the state’s development corporation or partner landlords are gifted revenue-generating assets free of oversight in perpetuity, gratis. Basically, at his most radical, Hemingway’s plan is to create a landlord state, wherein the state expands upon its existing mandate to merely preserve legal property rights and starts selling them.

At all points, Hemingway remains utterly unbothered by distributional questions and avoids entirely concerns over how the supposed magnanimity of capitalist endeavor (or social wealth in the form of housing) might accrue to the different classes. Instead of any details about distribution proposals, he instead shares an insipid anecdote about his own experience with a rent increase he begrudgingly accepted. Intended as an allegory about the paucity of tenant power, it is actually a parable in which his own petit-bourgeois spinelessness is elevated to a universal law: “I ended up ‘agreeing’ to an illegally high rent hike . . . so I could stay in my home.”

Why not attack landed property, and its administration by landlords, directly?

He further claims that this tells us that, in order “to secure renters’ rights, there needs to be enough housing supply so there are plenty of openings when individuals or families need or want to move.” But rental households cannot vote with their feet—try, as a tenant, to break a lease in order to move into a new, cheaper unit, and see how far you get before getting hauled into housing court. The existing supply is irrelevant at the moment of (self-)eviction because the tenant-landlord contract is inviolable on the part of the former while highly malleable for the latter. His insistence that an increase in supply proportionately increases tenant power is indirect, misleading, and dangerous. The real solution to his problem would be a direct increase in tenant power in terms of protections from rent increases, evictions, etc. It is stunning that he can look at his own experiences and not come to the conclusion that he should be protected from arbitrary increases, but instead should be able to literally flee from the problem. This leads one to wonder why, if providing tenants with a more even playing field vs. the landlords is his goal, he seems to think this is only possible via the most backward, circuitous route. Why not attack landed property, and its administration by landlords, directly?

“Pragmatic” progressives such as he cannot assault landed property because doing so opens up the dangerous possibility of an assault on all property. It’s not mine to say whether the worshippers of housing supply solutions, like Hemingway, believe themselves to be crusaders for the common tenantry or not—and frankly, it doesn’t matter. What is salient is that they look at the situation—in which rents stagger ever upward, in which units are systematically withheld from the market by the tens of thousands—and, instead of advocating for a decapitation strike on the responsible parties, propose a scheme of enrichment for capital accompanied by a false prophet’s promise of salvation. Stay the course, just a bit longer! Keep building, just a bit more!

If there is a long tradition of utopia, there is likewise one of economic utopias in the annals of capitalist history. In 1930, in the bleak spasm of the Depression, the economist John Maynard Keynes published an essay called “Economic Possibilities for Our Grandchildren,” wherein he famously proclaimed that “the economic problem may be solved, or be at least within sight of solution, within a hundred years” and that, by dint of the dizzying technological progress brought about by the capitalist mode of production, future generations would have to reckon with a world in which the “struggle for subsistence” was no longer the primary problem facing the human race. Keynes, at least, ensured that his New Jerusalem would be unrecognizable to his contemporaries; Hemingway can only look around at this fallen world and suggest the solution is to have more of it.

[*] Correction: A previous version of this article incorrectly cites a 2022 report from the Canadian Mortgage and Housing Corporation (CMHC) as calling for the construction of an additional five hundred seventy thousand residential units in British Columbia above a “business-as-usual scenario” per year by 2030 to restore affordability to levels last seen in the early 2000s. The report actually calls for that number to be constructed by 2030. We have corrected this—and the subsequent calculations—accordingly. We regret the error.