Simulating Impacts of Food Taxes and Subsidies

 Just last week I posted about food subsidies. I asked:

"Could you make the argument that simply shifting money toward programs related to fruits and vegetables would have a large enough impact on price to influence consumption? How much money would that take and what would the effect size be?"

This video discusses a PLOS medicine study that finds positive health benefits related to subsidies for fruits and vegetables and the referenced study (from a 'headline' reading) indicates a 10% subsidy could have a meaningful impact on health:

http://www.msn.com/en-us/video/tunedin/making-fruits-and-veggies-cheaper-could-save-a-lot-of-american-lives/vi-BBCQCE0

The video makes no direct reference to a study. However based on this reading I think the following are the studies mentioned in the video, the first being the primary reference:

Reducing US cardiovascular disease burden and disparities through national and targeted dietary policies: A modelling study. PLOS Medicine. June 6, 2017
https://doi.org/10.1371/journal.pmed.1002311

Taxes and Subsidies for Improving Diet and Population Health in Australia: A Cost-Effectiveness Modelling Study. PLOS Medicine. February 14, 2017
https://doi.org/10.1371/journal.pmed.1002232

Both papers are simulations based on assumptions related to elasticities and projected health outcomes (vs. empirical findings). One thing advocated is to create a revenue neutral policy that taxes soda and subsidizes fruits and vegetables. I'm skeptical, but if I want to give any credit I could say cynically that in this way maybe a soda tax could be useful even if it fails to have any meaningful impact on consumption or health.

See also: 

Are Soda Taxes Effective?
Four Big Questions about Farm Subsidies