Where the pavement ends: urban green space is becoming refuge and risk

On a Wednesday afternoon in July 2026, a woman settled her tarp beneath a stand of oaks on the edge of a mid-sized North American city, choosing the patch of grass that stayed dry longest after rain. By dusk the same canopy was being used by raccoons; the raccoons, in turn, were being watched by a public-health officer who had stopped counting clinic visits for rat bites after the third consecutive Wednesday. The scene is no longer unusual, and a review published in mid-July argues that treating it as exceptional is exactly the mistake policymakers keep making.

The point of departure is mundane. Homelessness is rising in most large economies, driven by housing costs that have outpaced wages for the better part of a decade. People who cannot afford a bed are pushed into the parks, riverbanks, highway verges and overgrown lots that still exist inside the urban footprint. Those same patches are the last continuous habitat for raccoons, coyotes, foxes, rats, snakes, geese and the bacterial communities they carry. The two populations now meet more often, for longer, and with less buffer than at any point in living memory.

What the overlap looks like

The new review, summarised on Phys.org on 15 July 2026, catalogues the recurring patterns reported by social workers, wildlife rehabilitators and epidemiologists across several countries. Unhoused encampments draw rodent activity because food waste accumulates where there are no bins and no regular collection. Raccoons and foxes habituate quickly to human presence when there is no threat of removal; the animals that survive long enough to breed produce offspring that treat the encampment as neutral ground. Snakes show up where cover is dense. Canada geese colonise the lawns where dogs are not walked. The author of the review, Danielle E. Mills of Trent University, frames the contact as bidirectional: the unhoused population is exposed to zoonotic disease and bites, and the urban wildlife population is exposed to concentrated human waste, deliberate harm, and the secondary poisoning that follows rodenticides laid down by private contractors.

The epidemiology is uneven but real. Rat-bite presentations to urban emergency departments have been climbing for several years in cities that track the figure; the review attributes the trend in part to encampments, in part to warmer winters that extend rodent breeding seasons. Rabies exposure in some jurisdictions has migrated from the rural fringe into parks where raccoon density is high. Avian influenza in urban geese, well documented across North America and Europe since 2022, adds another vector. None of these are novel pathogens; the novelty is the density and the persistence of the contact.

The reflex that makes it worse

The standard municipal response treats the wildlife as the problem to be managed and the encampments as the problem to be cleared, and tends to do both badly. Relocation of encampments fragments the social networks that encampment residents rely on for safety, pushes people into less visible sites where sanitation is worse, and does nothing to the rodent population that has already moved in. Culling or trapping raccoons and geese opens a niche that rats fill faster than the original species is removed. The two interventions interact: clear an encampment, and the rats that depended on it move into adjacent residential blocks; trap the rats, and the raccoons lose a competitor and breed more.

Mills's review, as summarised in Phys.org, argues for a combined approach that is closer to harm reduction than to pest control. Recommendations include secure, wildlife-resistant food-storage lockers co-located with sanitation facilities; routine waste collection at known encampment sites; vaccination and parasite-treatment programmes for free-roaming cats and dogs that act as bridges between wildlife and people; and a ban on second-generation anticoagulant rodenticides in proximity to encampments, on the grounds that secondary poisoning of raptors and foxes is now better documented than the public-health benefit. None of this is technically novel; what is new is the insistence that the two populations be planned for together rather than separately.

The structural frame

The pattern sits inside a larger story about who is allowed to be where. Housing markets in most OECD economies have moved faster than welfare systems: rents in nominal terms have roughly doubled in two decades, social-housing stock has shrunk in absolute terms in several countries, and the populations now sleeping in parks are not a fringe but a measurable share of the urban poor. Green-space access, in the meantime, has been mapped and remapped as a public-health asset: city planners sell parks as exercise infrastructure and as climate buffer, then discover that the same parks are being used as bedrooms. The two planning conversations rarely share a meeting.

The review is unusual only in naming the wildlife side of that collision. It notes, for instance, that urban-raccoon density in North American cities has roughly tripled since the 1990s, that coyotes have colonised every major Canadian city, and that fox populations in British towns have recovered to nineteenth-century levels after decades of decline. The animals are not invading; they are re-occupying habitat that has been depopulated of other species, including, increasingly, humans able to afford somewhere else to sleep.

Stakes and what to watch

The cost of doing nothing is paid first by the people in the encampments, who have the least access to medical care when bites and fevers occur, and secondarily by the wildlife rehabilitators and veterinary clinics that absorb the surge in injured animals. The cost of doing the wrong thing well is higher still: a 2018 study cited in the review estimated that routine rat poisoning in the United States kills more than a thousand raptors a year in treated areas alone, a figure that has not fallen since. The interest alignment is closer than the politics suggest: animal-welfare organisations, public-health agencies, and housing advocates all benefit from secure storage, regular waste collection, and more housing. The political alignment is harder, because each of those constituencies is funded and governed through different channels.

Two things are worth watching in the second half of 2026. First, a handful of mid-sized North American cities are running pilot programmes that bundle rodent-resistant waste infrastructure with outreach workers, and the early data, when it lands, will test whether Mills's combined model outperforms the relocation-and-trap approach on both counts. Second, the wildlife side of the ledger is poorly measured; the review's call for standardised monitoring of raccoon, rat and coyote density in and around encampments is, in effect, a request for the data infrastructure that would let cities plan for the overlap rather than react to it. Without that, every wet spring will bring another round of bite reports, and every wet spring will produce another set of recommendations that fail to mention who is sleeping under the trees.

Monexus framed this as a planning failure first and a wildlife story second, inverting the usual emphasis. The wire coverage on 15 July concentrated on the zoonotic risk; the structural point is that the risk is a symptom of a housing market that has stopped absorbing the people it used to.