This graph is just the overall death rates by age group in NYC and not IFR. However, you can make inferences from it. The death rate in 18-44 is roughly one tenth of the overall which would suggest around 0.1%. That assumes the infection rate is similar between overall and 18-44. I would guess that some of those in the 18-44 are not healthy and have some comorbs so I would guess the IFR for healthy people 18-44 would be some fraction of 0.1%.
Correct. And if one wants to dive into the numbers, the graph is deaths per 100,000 total population in each group - as a check on that, the overall 208/100K number x 8.5MM population in NYC = 17,680 NYC deaths through 6/18, which is correct (for deaths confirmed via PCR test; the presumed death total for NYC is 22,244). So, for 18-44 it's simply 0.02% (20/100K) and given NYC's ~20% infection rate, via seroprevalence antibody testing, the IFR for 18-44 would be about 5X that or 0.1% and yes, the IFR for healthy 18-44 year olds would be less than that.
If anyone is curious, while this sounds low, it's still far greater than the IFR for influenza in an average season, which is 0.02% for symptomatic flu cases (in the 18-49 age group the CDC has data for), but likely about 0.01% on a total infection fatality rate, including asymptomatics (estimates for this are all over the place though, like 20-80%, so I took 50%), which is 1/10th of that for coronavirus, roughly, which makes sense. Right now we're seeing 1.0-1.2% IFR in places like NY/Spain that actually have seroprevalence data for their entire population (which would be about 1.5% vs. symptomatic cases assuming 65% of CV infections are symptomatic) and in a typical year influenza has a symptomatic IFR of 0.1% overall.
https://www1.nyc.gov/site/doh/covid/covid-19-data.page
https://www.cdc.gov/flu/about/burden/past-seasons.html
https://www1.nyc.gov/site/doh/covid/covid-19-data.page