The SPIQ-FS dataset of
marginal damage costs
The SPIQ-FS dataset of marginal damage costs was developed as background input to the Food System Economic Commission.
The SPIQ-FS generic dataset version 0 contains national estimates of marginal damage costs in 2020 USD PPP for the emission or production in 2020 of 14 impact quantities in 158 countries. The quantities are listed below. Uncertainty in marginal damage costs are reported by parametric probability distributions. Correlation matrices are reported in the SPIQ-FS dataset, allowing assessment of economic risk for joint changes in quantities.
The intended use involves aggregation across countries and quantities. Examples include global studies of dietary change, reporting the impacts of the annual operations of a multinational food or agriculture company, or estimates of economic impact along global supply chains.
The marginal damage cost estimates should not be used for local or site-specific studies.
SPIQ-FS marginal damage costs are measured in 2020 USD PPP (Purchasing Power Parity), also known as 2020 International Dollars. Purchasing power parity represents the equivalent amount of a basic goods basket in 2020 that $1 USD, once exchanged to local currency, purchases in that country. The goods represent welfare provided by their consumption. Damage costs measured in 2020 USD PPP represent the reduction in welfare due to reduced purchasing power and avoided damage costs represents the benefit in an avoided reduction in welfare.
Download the generic SPIQ-FS dataset or download the SPIQ-FS documentation.
Quantities associated to
food system impacts
Quantities associated to external costs and market failures of the food system in the generic SPIQ-FS dataset.
| Environmental | Quantity | Unit | Notes |
|---|---|---|---|
| GHG emissions | C2O to atmosphere | metric ton | |
| GHG emissions | CH4 to atmosphere | metric ton | |
| GHG emissions | N2O to atmosphere | metric ton | |
| Nitrogen emissions | NH3 to atmosphere | kg N-weight | |
| Nitrogen emissions | NOx to atmosphere | kg N-weight | |
| Nitrogen emissions | Nr to surface waters | kg N-weight | |
| Nitrogen emissions | NO3- to groundwater | kg N-weight | |
| Water use | Blue water withdrawal | cubic metre | |
| Land use changes | Forest habit lost | ha | Loss of ecosystem services from intact forest |
| Land use changes | Forest habitat returned | ha | Return of ecosystem services in forest regenerating from use as cropland or pasture |
| Land use changes | Other land habitat lost | ha | Loss of ecosystem services from intact woodland, shrubland, and grassland. |
| Land use changes | Other land habitat returned | ha | Return of ecosystem services in woodland, shrubland, and grassland regenerating from use as cropland or pasture |
| Diets | Quantity | Unit | Notes |
|---|---|---|---|
| Undernourishment | Number of undernourished (FAO) | person | Productivity loss from undernourishment as defined by the FAO |
| Non-communicable diseases | Food supply (FAO) | metric ton | Coming soon in SPIQ version 1 |
| Obesity | Number of overweight and obese | person | Coming soon in SPIQ version 1 |
Use for economic cost
The average present value of probable 2020 USD PPP marginal economic loss given a unit change in a quantity is given in the SPIQ-FS dataset. The average value should be used to calculate the average value of total economic losses across multiple countries and joint changes in quantities since it is additive.
To calculate risk in total economic losses across multiple countries and joint quantity changes the correlation matrices in the SPIQ-FS dataset should be used to reconstruct a joint distribution of probable 2020 USD PPP present values for the impact quantities. Samples taken from the joint distribution of marginal damages should be multiplied by their respective quantities for each country and then added. The resulting set is a sample of total economic losses. Economic risk can be underestimated without using joint sampling.
It is not recommended to use the average values in the SPIQ-FS dataset separate from the uncertainty estimates.
Use for economic cost-benefit
The marginal damage costs in the SPIQ-FS dataset and any totals for economic losses calculated using them do not include the benefits provided to society from production of the respective impact quantities.
For example, the damage costs for a metric ton of ammonia (NH3) emissions to air following application of synthetic fertliser on a field are calculated by multiplying the NH3 to atmosphere marginal damage cost in SPIQ-FS by the weight of nitrogen in kg of a metric ton of ammonia (822 kg). The damage cost does not include an adjustment for the agricultural value add produced by the application of the synthetic fertiliser.
Damage costs should be paired with abatement costs and counterfactuals to determine the economic risk from food system activities, and the economic potential of transition to a lower impact food system. Abatement costs include the option of ‘paying the cost’ of losing the value from production or consumption, but often there are lower cost options to reduce quantities but retain the same value from production or consumption.
There is no indication without a counterfactual study that the economic losses represented by the damage costs are recoverable, or that it is even desirable to do so if the abatement costs are higher than the value of abatement represented by reducing or avoided damages.
The simplest use with a counterfactual is two or more scenarios which, all else being equal, have the same value from production or consumption with differences in impact quantities. In this case, a change in overall economic value is estimated by the marginal damage costs multiplied by the change in the respective impact quantities.
SPIQ-FS Datasets
SPIQ-FS datasets are generated from costing models with common modules. Components of the models, such as social discount rates, value of ecosystem services in aquatic and terrestrial biomes, and the productivity costs of a disability adjusted life year lost to poor health from a disease outcome such as type II diabetes can be customised for individual studies.
A SPIQ-FS dataset is built by the following process
- Build of marginal damage costs from country data and the quantities and counterfactuals to be studied. Customisation of costing components such as social discount rate parameters and GDP growth rates for countries is included in build files.
- Import of quantity changes from economic modelling which may include categories such as quantity of CH4 emissions from enteric fermentation. During import quantities are adjusted to match the scale and unit of the built marginal damage cost.
- Calculation of a joint sample of damage costs for individual cost items. A cost item is a paired quantity and marginal damage cost with the same scale and unit.
A global study involving 160 countries, 2 counterfactuals and 30 categories of quantity changes will have 9600 individual cost items and a joint sample of 9600 random variables representing the damage cost and its uncertainty for the cost items.