This step seeks to project future biodiversity expenditures based on historical trends. Future projections should cover a forward period of approximately 5 to 10 years. The exact time period chosen will depend on national budgeting processes and cycles, identified in the PIR (Chapter 3). The projections should cover the timing of the national biodiversity strategy at a minimum.
Clearly documenting and validating methodological decisions and assumptions used with stakeholders is essential. Where the trend does not depict erratic behaviour, and history is our best teacher, a long-term average growth rate can be applied as a factor. A linear regression analysis reveals the average level of budget expenditure and the annual rate of change over the time period and facilitate extrapolation into the future. It also provides a measure of variability around the average that might be used for sensitivity analysis, ‘optimistic’ and ‘pessimistic’ projections.
Sensitivity analysis is typically conducted by changing several key variables and assumptions in projected expenditures to identify those assumptions that may be the most impactful if changed. Where greater precision in estimates and predictions can be secured, we can expect less sensitivity to change.
When we have information about past budgets that leads us to believe that their predictive power is questionable, we can follow alternative paths. We can accept expert opinion or an ad hoc algorithm to combine what we know about the future with what we observe in the past; for example, moving average or trend analysis based on biodiversity expenditures as a percentage of government budget or GDP. Countries that historically demonstrate a high degree of variation in rates of exchange and inflation may want to adopt a threeor five-year weighted moving average (WMA) approach to reduce prediction error in future forecasts due to unusual or temporary short-term variations in these factors. However, the effectiveness of the models depends on the quality of the data.
Examples of projections are profiled in Annex 1, including: