From global biodiversity commitments to local action: revenue potential and allocation dynamics of the Cali Fund

Established at CBD COP16 in 2024, the Cali Fund is the first global mechanism requiring private companies benefiting from genetic data to share 1% of their profits or 0.1% of revenues, with biodiversity-rich countries and local communities. I provide the first global firm-level estimates of the fund’s revenue potential and a systematic analysis of its distributional properties on both the contribution and allocation sides under alternative design choices, using firm-level data for 21,690 eligible companies across 100 countries. Under realistic voluntary uptake, annual contributions would reach approximately $0.9 billion, rising to $3.6 billion under full participation. The revenue base is narrow: two sectors account for 51% of contributions under full compliance and the top 100 contributors for 27.6%. Under realistic voluntary uptake, sectoral composition shifts further, with pharmaceuticals alone accounting for nearly 50% of expected revenues under the central scenario. For most sectors the payment rule ties contributions directly to profitability, making fund revenues sensitive to economic conditions: firms clustered near the zero-profit threshold can move in and out of contribution obligations as profits fluctuate. Pharmaceuticals is the exception: most eligible firms in this sector pay on revenue regardless of profit swings, providing a stable base but making overall fund revenues heavily dependent on pharmaceutical participation. On the recipient side, the distribution of who receives how much is largely determined by the way the allocation formula is designed. Prioritising the geographic origin of genetic data channels nearly half of all resources to wealthy research economies while African States receive 6.4%. The mathematical form of the formula also matters: the formula and normalisation method together can shift allocations by up to 13 percentage points away from Western countries without changing any stated weight. With both contribution rules and allocation criteria still open for revision at COP17 in October 2026, the paper shows these choices are not distributionally neutral. ications in which parameters are interpreted cautiously as reduced-form components rather than direct measures of underlying behavioural channels. The dissertation first revisits Lotka’s systems perspective and uses it to organise DEMs relevant to demography within a common formulation that separates population state variables from time-varying characteristics or environmental conditions. DEMs are grouped into basic growth, dynamic heterogeneity, and structural heterogeneity families, and a practical workflow is outlined: initial conditions, analytic versus numerical solutions, and parameter estimation by non-linear least squares supported by global optimisation. It also discusses why DEMs have seen limited development in demography and summarises their strengths, limitations, and potential applications. The empirical chapters develop and apply two fertility models using cohort schedules from the Human Fertility Database. Chapter 4 models cohort first-birth timing by extending Bass diffusion model and introducing a simplified two-type representation of latent heterogeneity under outcome observability. Chapter 5 extends the framework to birth-order dynamics by jointly modelling multiple-birth schedules. The specification embeds parity progression, uses neighbouring cohorts to construct reference-set indices, and includes a joint exposure term. Overall, the dissertation shows how DEMs can be implemented with standard cohort schedules to provide compact summaries of fertility timing and sequencing, while highlighting directions for future work on measurement and social interaction structure.