A few items of note, mostly from the new batch of work put out through the National Bureau of Economic Research:
- There’s an Australian paper from April 8 that’s making the rounds thanks to a boost from Tyler Cowen. Basically, it looks at the death predictions made by the University of Washington coronavirus model, compares them with what actually happened, and finds them wanting. It focuses on the state-level predictions for dates through April 2, so it doesn’t reflect the big revisions made to the model in the past week or so, but what’s striking is that most of the death totals wound up outside the uncertainty intervals the model provides. (When your 95 percent interval includes the correct number only 30 percent of the time, that’s kind of a problem.) The paper does not find that the predictions were consistently off in a specific direction, notably, but this certainly gives policymakers another reason to be skeptical of the model. As I wrote before, look at what it says and take it into consideration, but also consider other evidence and bear in mind that no one really knows what’s going to happen.
- The New England Journal of Medicine has a fascinating finding from a New York hospital that tested 214 women giving birth: 33 tested positive (about 15 percent), of whom only seven showed or developed symptoms. Others might yet develop symptoms, there might be some false positives in there, and this isn’t exactly a random sample of New Yorkers, but this is at least suggestive evidence of a lot of asymptomatic cases that usually escape detection.
- State lockdown orders seem to work, according to a study that basically compares the timing of the orders with later patterns in COVID-19 growth rates. “A hypothetical national stay-at-home order issued on March 13, 2020 when a national emergency was declared might have reduced cumulative county infections by 62.3%, and might have helped to reverse exponential growth in the disease by April 5.”
- A survey finds that “43 percent of [small] businesses are temporarily closed, and businesses have — on average — reduced their employee counts by 40 percent relative to January.” These businesses are economically fragile, have no idea how long all this will last, and hope they can get relief funds.
- The economy started to contract the week ending March 21, before most of the official lockdowns.
- One study arrives at a “conservative” estimate of “a cumulative loss in industrial production of 12.75% and in service sector employment of nearly 17% or 24 million jobs over a period of ten months.”
- Another: “Our illustrative exercise implies a year-on-year contraction in U.S. real GDP of nearly 11 percent as of 2020 Q4, with a 90 percent confidence interval extending to a nearly 20 percent contraction.”
- George Borjas takes a look at the demographics of testing and infection rates in New York.
- Yet another study, this one trying to figure out the best lockdown schedule: “The optimal policy prescribes a severe lockdown beginning two weeks after the outbreak, covers 60% of the population after a month, and is gradually withdrawn covering 20% of the population after 3 months. The intensity of the lockdown depends on the gradient of the fatality rate as a function of the infected, and on the assumed value of a statistical life.”
- With existing data, it’s really hard to estimate the disease’s fatality rate.