Inconsistent reporting of variable "Policy Cost|Consumption Loss"

[nroming]

For some mode/scenario combinations, this seems to be computed based on discounted, for other on undiscounted consumption losses relative to the given baseline scenario.

[danielhuppmann]

Thank you @nroming for identifying this issue.

This variable is included as reported by the modelling team, and the variable definition (see below) is actually not explicit on how to compute the value.

Policy Cost|Consumption Loss:
consumption loss in a policy scenario compared to the corresponding baseline (losses should be reported as negative numbers) (billion US$2010/yr)

I'm not sure how to fix this issue without re-collecting all variables from all modelling teams.

[nroming]

One could add a new variable ("Harmonized Policy Cost"?) (or even replace the existing one) applying the same methodology based on discounted consumption pathways.

A short term solution would be to just drop it?

[danielhuppmann]

I would be cautious about removing any data.

If someone (nudge, @nroming?) would feel up to writing a script that computes a new harmonised policy cost timeseries, we could include this information in the next data release.

[nicobauer]

Hi Daniel, Niklas,

unfortunately I find even more issues with the "policy cost|*" variables. I am currently browsing through these variables.

Regarding Niklas's proposal. It is not so many modeling teams that have submitted policy costs. It should be possible to make a request.

For the case of REMIND: we report current values (i.e. not discounted).

Best
Nico