It might be worth widening your distribution for the effects on supply/demand further into the negative and checking sensitivity to this value and the size of the welfare improvement for individuals.
On further consideration, I’d imagine the effects of the welfare improvement will dominate the effects of change in number (as they do now), since they’re a few orders of magnitude larger in absolute value, anyway.
For dissolved oxygen in Taiwan, it was 30 billion welfare points from the welfare improvement and 610 million welfare points from reduced supply/demand.
For feed fortification in India, 180 million welfare points from the welfare improvement and 60,000 welfare points from reduced supply/demand.
So flipping the signs of the supply/demand effect wouldn’t make much difference.
It might be worth widening your distribution for the effects on supply/demand further into the negative and checking sensitivity to this value and the size of the welfare improvement for individuals.
On further consideration, I’d imagine the effects of the welfare improvement will dominate the effects of change in number (as they do now), since they’re a few orders of magnitude larger in absolute value, anyway.
For dissolved oxygen in Taiwan, it was 30 billion welfare points from the welfare improvement and 610 million welfare points from reduced supply/demand.
For feed fortification in India, 180 million welfare points from the welfare improvement and 60,000 welfare points from reduced supply/demand.
So flipping the signs of the supply/demand effect wouldn’t make much difference.