The scale of the problem is not in dispute. Empirical audits in Canada found that 62.5 percent of fish habitat compensation projects resulted in a net loss of habitat productivity, and that 82 percent were out of compliance with their authorizations. In Quebec, inadequate application of the mitigation hierarchy was linked to the loss of 99 percent of legally disturbed wetland habitat between 2006 and 2010. A separate analysis found that 67 percent of reviewed Canadian projects were authorized to impact more habitat than they were required to compensate for, while a recent assessment of offsetting for species at risk in Ontario found that benefit criteria were not consistently met and that completion timelines were often missing. Similar shortcomings have been documented globally, underscoring what the authors describe as an urgent need for more transparent and enforceable offset frameworks.

Offsets can fail for several interacting reasons, including poor equivalency between what is destroyed and what is restored, non-compliance with permit conditions, time delays before restored habitat becomes functional, and simple uncertainty about whether restoration will work at all. Field assessments themselves introduce noise: one recent study reported an average variation of roughly 34 percent in offset credits depending on who conducted the assessment and how it was performed. Monitoring is frequently inadequate or poorly designed, leading to overly optimistic assumptions about offset effectiveness. Multipliers, which inflate the size of an offset to buffer against these failures, have long been advocated as a solution, but few jurisdictions prescribe how to compute them. Canada’s own aquatic offset policy states that time lags and uncertainty must be accounted for, yet provides no methodology for doing so, leaving managers without a defensible basis for requiring larger offsets.

The new framework, developed by Adam S. van der Lee, Madison E. Brook, and Marten A. Koops, builds on an earlier approach proposed by Michael Bradford in 2017 but extends it to complex projects involving multiple impacts and multiple offset sites. The core idea is conceptually simple. Instead of treating the value of an impact and the value of an offset as fixed numbers, both are represented as probability distributions. The researchers chose lognormal distributions, which are continuous, always positive, and become increasingly right-skewed as uncertainty grows, a property they argue better reflects the real range of ecological outcomes than the normal distributions used previously. The ratio of the impact distribution to the offset distribution then yields an entire distribution of possible multipliers, from which a single value is extracted according to a management-defined risk tolerance.

Risk tolerance is the framework’s central lever, and the authors show how it translates directly into compensation requirements. At a risk tolerance of 20 percent, the 80th percentile of the multiplier distribution is selected, accepting a one-in-five chance that offsetting ends in a net loss. Tightening the tolerance to 10 or 5 percent raises the multiplier, improving the odds of achieving no net loss or even net gain. Previous work has suggested multipliers might need to exceed 100 to guarantee success, while others contend that no net loss may be unattainable regardless of multiplier size. Yet there is also evidence that multipliers below 5 can suffice in freshwater ecosystems. The simulations conducted by the Canadian team, which involved 100,000 replicate draws for each scenario, landed squarely in the modest range: uncertainty multipliers spanned from 1.05 under favorable conditions to 6.98 in the most demanding scenario, and they consistently delivered no net loss at the designated risk tolerance across every tested combination of offset count, uncertainty level, and covariance.

One of the framework’s most consequential innovations is a weighting parameter that allows offset-specific multipliers rather than a single blanket multiplier applied across all compensation sites. Large development projects routinely involve several offset locations; one review of French projects found an average of 3.8 offset sites per development. Because expanding a particular offset may be cost-prohibitive, logistically impractical, or ecologically undesirable, the model lets planners assign each offset a weight between zero and one reflecting its capacity to absorb additional compensation. The simulations revealed that this choice matters enormously. When two offsets carried unequal uncertainty, applying the multiplier only to the more uncertain offset required more than twice the total compensation compared with directing it to the more reliable one, with summed multipliers of 7.22 versus 3.39 under a 5 percent risk tolerance.

The mathematics also illuminate a subtle and counterintuitive role for covariance, the degree to which the fates of impacts and offsets are statistically linked. When impacts and offsets share environmental drivers, for example because they sit in the same watershed and respond to the same climatic fluctuations, high covariance actually reduces the required multiplier, since random deviations in losses and gains tend to cancel out. Under perfectly correlated conditions with equal uncertainty, the multiplier collapses to exactly one. But when offset uncertainty greatly exceeds impact uncertainty and covariance is high, the required multiplier climbs substantially, becoming 1.6 to 1.74 times larger than under independence, because the offset’s random swings systematically dwarf those of the impact. Given that strong covariance is unlikely and difficult to quantify in practice, the authors recommend assuming independence unless empirical evidence supports otherwise, a precautionary default that avoids embedding poorly justified assumptions into regulatory decisions.

To demonstrate the framework’s practical value, the authors worked through a hypothetical development project destroying five hectares of high-quality riparian habitat, to be compensated by a mix of riparian restoration and construction of off-channel habitat such as a pond. Under a 10 percent risk tolerance, treating both offsets equally would demand 13.78 hectares in total at an estimated cost of 7.03 million dollars. Relying solely on riparian restoration would require 55.7 hectares at 9.47 million dollars, while relying solely on habitat creation would need 10.31 hectares at 8.76 million dollars. By iteratively adjusting the weighting parameters, the framework identified a cost-minimizing solution: applying a multiplier of 14.4 to the restoration project and 4.9 to the habitat creation project, yielding 19.3 hectares in total at 6.58 million dollars, roughly 450,000 dollars cheaper than the equal-weighting strategy. The example illustrates how mixing offset types can outperform any single-offset approach while still meeting ecological risk thresholds.

The authors are candid about the framework’s limits. No multiplier can rescue an offset that fails completely, and the method assumes both regulators and proponents act in good faith toward the shared goal of no net loss; it cannot guard against negligence or deliberate underestimation of impacts. It also does not address permanence, since offsets can deteriorate over time without maintenance and adaptive management, and it treats uncertainty separately from time delays, which must be handled through an independent time-delay multiplier derived by comparing the schedule of impacts against the schedule of offset implementation, ideally over a time horizon twice the project duration. Nonetheless, the researchers argue the approach offers something offset policy has lacked for decades: a transparent, flexible, and statistically grounded standard that gives regulators a defensible basis for enforcement, gives proponents cost certainty and faster approvals under the polluter-pays principle, and embeds bet-hedging into offset design by spreading risk across multiple sites. The team has released an interactive Shiny application and R code so that practitioners anywhere can begin applying the method, potentially transforming biodiversity offsetting from a matter of negotiation into a matter of calculation.

Subject of Research: A quantitative Monte Carlo framework for calculating biodiversity offset multipliers that incorporate uncertainty and risk tolerance

Article Title: A Quantitative Approach to Biodiversity Offset Multipliers: Managing Uncertainty in Complex Projects

Article References: van der Lee, A. S., Brook, M. E., & Koops, M. A. (2026). A Quantitative Approach to Biodiversity Offset Multipliers: Managing Uncertainty in Complex Projects. Environmental Management, 76(9), Article 311. https://doi.org/10.1007/s00267-026-02623-w

Image Credits: AI Generated

DOI: 10.1007/s00267-026-02623-w

Keywords: biodiversity offsetting, no net loss, offset multipliers, Monte Carlo simulation, environmental management, ecological uncertainty, risk tolerance, habitat compensation, restoration ecology, conservation policy, lognormal distributions, regulatory defensibility