More on that serology study I noted below. My general point stands about how vulnerable and relatively untouched the California population remains based on these numbers. But there’s another key point to take note of. TPM Reader BR focused my attention on this. If we extrapolate the total number of infections across California, this shows just how massively the official infection numbers undercount what appears to be the actual number of California residents who’ve been infected.
Some basic numbers.
The Santa Clara study makes a weighted estimate that 2.8% of people were infected and developed antibodies. The actual number of the people they tested who came back positive was 1.5%. They weighted this by demographic characteristics and depending on the assumptions got a range from 2.49% to 4.16%. For these purposes we’ll stick to the 2.8% ‘average’.
The population of California by the 2019 Census estimate was 39,512,223. Given how diseases spread you can’t just take the sample from Santa Clara and extrapolate it to the whole state which is larger than most countries and has several distinct major population centers across its geography. The study rightly notes that this study can only be considered representative for Santa Clara County. But there are multiple hotspots throughout the state. And here I am extrapolating this percentage statewide only to give a sense of the scale of the likely difference between lab confirmed infections and actual infections.
2.8% of California’s population is roughly 1.1 million people. The current number of lab-confirmed positives is 26,182. At the time this study was conducted (April 3rd and 4th) there were 12,026 cases. So over 90 actual infections for each lab confirmed one.
To ground this a bit more in apples to apples comparisons the study suggests that between 48,000 and 81,000 were infected with Coronavirus at a time (April 3rd and 4th) when there were 12,026 cases statewide at the time. (The state population is about 20 times the size of Santa Clara County.
I would risk the wrath of the statistical goods if I don’t remind you that we’re throwing together a lot of variables here, each of which contain significant uncertainties. And on top of that we’re making one significant apples to oranges leap. So keep that in mind. But this is some significant indication that the amount of true infections is likely vastly larger than the lab-confirmed numbers we have.
