--- title: why conservation projects fail on their own canonical_url: https://ensurance.app/guide/why-conservation-projects-fail-on-their-own markdown_url: https://ensurance.app/guide/why-conservation-projects-fail-on-their-own.md subtitle: the pilot that stands alone is the pilot that ends. the work was good. the structure was not category: nature-finance --- # why conservation projects fail on their own *the pilot that stands alone is the pilot that ends. the work was good. the structure was not* The project worked. The crew knew the ground, the monitoring came back clean, the partners showed up for the site visit, and the final report said what you hoped it would. Then the grant ended, and the question nobody put in the proposal arrived on schedule: who pays for year four? If you searched **why conservation projects fail**, you are probably not asking why the work was bad. You are asking why good work stops. That is the honest definition of failure here: a project that did something real and then had no way to keep doing it. The fence line goes unwalked, the monitoring lapses, the invasives find the gap, and the acres that were restored begin, slowly, to un-restore. The headwater forest, the floodplain, the meadow, and the run of fish exist — and already trade water, shade, soil, and food with each other — whether or not anyone funds a project on them. Ensurance gives each one an account so the people who depend on it can pay into the network. The network is not the forest. This post is about why a project asked to stand by itself tends to end, and why the fix is not a bigger grant. :::johnson **the pilot that stands alone is the pilot that ends.** Not because the work was wrong — because nothing was built to carry it once the first funder stepped away. [how nature already runs a portfolio →](/guide/nature-already-runs-a-portfolio?from=guide) ::: ## failure means the work stops A conservation project is a bounded effort on a bounded place: a reach of river, a stand of trees, a stretch of fence, a few thousand acres of fuel treatment. It has a start date, a budget, a deliverable, and an end date. **A conservation project fails when the living work it started cannot be continued, not when the work itself was poor.** The culverts were replaced. The willows took. The fish came back. Then the project ended, because projects end, and the place kept needing what it had been providing. The people who ran it usually did everything right. The failure is structural. It belongs to the shape of the thing — one funder, one site, one output, no neighbors — not to the people who stood inside that shape and did good work anyway. ## four single points of failure A standalone project can carry all four at once. Any one can end the work. ### one funder The grant is the cliff. A project funded by one source — a foundation program, an agency line item, a single corporate commitment — lives on that source's calendar, not the forest's. Programs change priorities. Agencies change administrations. Corporate budgets move with the quarter. The cycle ends and the project has two options: find another single funder and begin the cliff again, or stop. In the last year of a grant, staff time drifts toward the first option, and that is time not spent on the ground. What is lost is not just money. It is continuity: the crew that knew the site disperses, relationships with neighboring landowners cool, and the next funder, if one comes, pays to rebuild knowledge the project already had. ### one site A project on one place is exposed to everything that can happen to one place. One fire season. One flood that resets the channel. One sale to a new owner with no interest in the easement's intent. Nothing was done wrong; all of the project's value simply sat in one place. California's forest offset program shows this shape at scale. The program built a shared reserve — a buffer pool — into which every project contributed credits to cover reversals like fire. Badgley and colleagues found that wildfires depleted nearly one-fifth of the total buffer pool in less than a decade, equivalent to at least 95 percent of the credits set aside to manage fire risk for 100 years (*Frontiers in Forests and Global Change*, 2022). Read that carefully, because it is not a verdict on offsets. The buffer pool was an honest attempt to spread single-site risk across many sites. What the number says is that when each project is a single stand one fire can take, a thin shared reserve of such stands does not hold against the fire seasons that actually arrive. It is a fact about single-site exposure, and the developers who built those projects are cousins in this work, not villains in it. [A carbon credit is not a forest](/guide/a-carbon-credit-is-not-a-forest?from=guide) covers what the credit does and does not prove; this post is only about the place behind the credit standing alone. ### one output A project usually gets funded for one thing. A tonne of carbon. A credit. An acre-foot of water. A mile of fence. The funder's accounting needs a unit, so the project is sold as that unit. But the place does many things. The same treated forest holds soil on the slope above the county road, slows the snowmelt that fills the utility's reservoir, shades the creek the trout live in, and carries the elk between winter and summer range. Sold as one output to one buyer, every other output is given away — and everyone else who depends on the place stays an audience rather than a payor. That is the quiet cost of one output. Not that the tonne was fake, but that the forest was doing many jobs and only one had a customer. ### no neighbors The fourth failure connects the other three. A standalone project is not paid by the people downstream who depend on what it produces, and it does not pay into the projects upstream it depends on. The meadow restoration holds water the ranch below uses in August. The fuel treatment above the reservoir keeps ash out of the intake. The riparian fencing on one parcel only matters if the parcel upstream is not loading the creek with sediment. Each depends on the others, and each is funded as if it stood alone. So when one ends, the ones it was quietly supporting lose ground too, and no ledger shows it. The forest, the floodplain, the meadow, and the fish were already trading with each other. The money never was. ## the work was good Before the turn, the part that needs to be said plainly: standalone projects have restored real acres, real reaches, real populations. The land trust director who has run one well for twelve years did not make a mistake. The watershed council that strung grants together to keep a crew in the field did exactly what the system allowed. The critique here is of the object — a project asked to carry itself — not of the people who made it work for as long as it could. How long these projects last is evidence of how much skill it takes to hold a lonely structure upright. The network does not replace that work. It is what that work joins. ## the fix is not a bigger grant When a good project is about to end, the instinct is to find a bigger funder for the same project. Sometimes that buys another cycle. It does not change the shape. A larger single grant is still one funder, one site, one output, no neighbors — just a taller cliff. The fix is to change what holds the project up. First, the people who already depend on what the project produces pay into it. Not "the public," not "donors who care about nature" — the utility whose intake the forest protects, the county whose road the slope holds, the people who fish the reach. [People don't care about nature; they care about theirs](/guide/people-dont-care-about-nature-they-care-about-theirs?from=guide) makes that case in full. Here it is enough to say that a payor who depends on the place has a reason to stay when the grant ends, because the dependency does not end with it. Second, the project routes a share of what it receives to the projects it depends on — the one upstream, the one on the slope, the one holding the corridor open. A project with payors below it and projects it pays above it is no longer a single point of failure. When one has a bad year, the others can still be paid, and can still pay. Alone, a project is one chance at one place. Linked, it is one node in something that can carry a new project while it finds its footing and hold an older one through a hard season. That is financial resilience and regional resilience at once, and the next post takes up how nature itself already runs that way. A short gloss, now that the idea stands on its own: each place or purpose can hold an account in its own name; a certificate is a share of one named place; proceeds are the routing that can send a set share of what one account receives to the accounts it depends on. The operator sets the split. [Projects that fund each other](/guide/projects-that-fund-each-other?from=guide) walks through the plumbing; [proceeds](/proceeds?from=guide) shows the live routing. The volumes are small today and the network grows one node at a time. The structure is what is different. ## frequently asked questions ### why do conservation projects fail? A good conservation project fails when its structure gives out, not its science: the work stops when the single funder's cycle ends, when one event hits the single site, when the place is sold as one output to one buyer, or when the project has no neighbors paying into it or being paid by it. The fix is to link it to the people who depend on it and the projects it depends on. ### what happens to a project when its only funder leaves? The work stops or goes into survival mode. Staff time shifts from the ground to the next proposal, the crew disperses, monitoring lapses, and the gains begin to erode. A project with several payors who depend on its output faces a lower cliff: their need does not end with the grant, and one payor leaving does not take all the funding with it. ### can a single project survive one bad fire season? Sometimes, but it should not have to bet on it. A project whose whole value sits on one site can lose it in one season, and California's forest offset buffer pool showed that a thin shared reserve of such sites did not hold. A project linked to others across places and link types can be carried through a bad year by the nodes that were spared. A season that burns across the whole region is still a hard year for every node in it; the network spreads the risk, it does not erase it. ## where to go next The next post shows that nature already runs this structure — several hundred salmon populations keeping one fishery open — and what that means for a network of projects: [nature already runs a portfolio](/guide/nature-already-runs-a-portfolio?from=guide). To see the places and purposes that already hold accounts, start at [explore](/explore?from=guide). ## sources [Badgley et al., "California's forest carbon offsets buffer pool is severely undercapitalized," *Frontiers in Forests and Global Change* (2022)](https://www.frontiersin.org/journals/forests-and-global-change/articles/10.3389/ffgc.2022.930426/full) — wildfires depleted nearly one-fifth of the total buffer pool in less than a decade, equivalent to at least 95 percent of the credits set aside for 100 years of fire risk [Reuters, "Wildfires are destroying California's forest carbon credit reserves," 5 Aug 2022](https://www.reuters.com/world/us/wildfires-are-destroying-californias-forest-carbon-credit-reserves-study-2022-08-05/) — news report on the study ## the series 1. [what a networked project actually is](/guide/what-a-networked-project-actually-is?from=guide) 2. [why conservation projects fail on their own](/guide/why-conservation-projects-fail-on-their-own?from=guide) — this post 3. [nature already runs a portfolio](/guide/nature-already-runs-a-portfolio?from=guide) 4. [landscape-scale conservation needs a way to pay across the fence](/guide/landscape-scale-conservation-needs-a-way-to-pay-across-the-fence?from=guide) 5. [how to fund the project you depend on](/guide/how-to-fund-the-project-you-depend-on?from=guide)