In its fourth week of sheltering-in-place, San Francisco’s rate of confirmed COVID-19 cases is growing at roughly half the rate of the state of California’s, and new data is showing promise of below-exponential growth.

“You can see the trend is going down,” said Dr. Nicholas Jewell, professor of biostatistics at UC Berkeley.

Jewell said that the average daily growth between March 26 and April 1 was 10 percent for San Francisco and 18 percent for California.

“Without being overly quantitative, it looks like San Francisco is increasing at half the rate that California is as a whole, and that is a big difference,” Jewell said.

He credited San Francisco’s lower growth rate to the shelter-in-place order that began March 17.

On March 18 and 19, he said, “you’re still seeing double-digit growth pretty much every day,” and on March 27, it begins to drop off. “That reflects a nine-day gap before you start to see the benefit of the social distancing, which is plausibly in the right range, which I would’ve said was a week,” he explained.

After hitting 25 percent daily rate increases on March 26 and 27 — some 10 days after shelter-in-place took effect— the daily rate of confirmed cases has not gone above 10 percent in almost two weeks.

Testing also ramped up from about 126 people tested in the first week of shelter-in-place to around 300 people tested per week in the succeeding two weeks, according to the Department of Public Health’s new data tracker unveiled on Tuesday.

Jewell said that more tests and less positive results is “a bigger evidence of the effect of social distancing.”

“We’re not necessarily testing everybody with symptoms,” said Dr. Rajiv Bhatia, Assistant Clinical Professor of Medicine at the Stanford School of Medicine.

“But the fact that the percentage of positives are remaining about the same is somewhat also consistent with the stability of new cases over the past two weeks,” he said. “That’s very positive. I think it’s very positive.”

“The hospitalizations occur perhaps on average about two weeks after exposure at most, about a week after symptoms develop,” he said. “So this tells me that there’s a stable growth in the number of cases about two weeks ago.”

Although data from San Francisco does not include the number of new admissions and discharges daily, Bhatia said that the net growth of five hospital admissions every day over the past two weeks is “reassuring.”

“That’s not growing exponentially,” Bhatia said. “I would feel comfortable saying that after shelter-in-place was implemented, we avoided exponential growth after that. We still have new cases, new transmissions but it is at a moderating pace.”

Dr. George Rutherford, Professor of Epidemiology and Biostatistics at UCSF, agreed that there were some hopeful signs. “We also have not crippled our hospital system.”

Rutherford compared a case wherein UCSF had seven patients on ventilators on Thursday while Columbia Presbyterian Hospital in New York City had 700.

“That’s 100 times more,” he said. “That easily could’ve been us without this early shelter-in-place order.”

“It may be a little early to declare victory and go for a victory lap around the track but the consequences of shelter-in-place are profound,” Rutherford said. “It’s thousands of deaths averted across the region.”

Peter Khoury, a data scientist who created the Phoenix Data Project that localizes COVID-19 data, said that San Francisco has a “head room now” with managing a surge.

He said that compared to the state of California or the whole United States that reaches exponential growth about every six days, San Francisco only grows exponentially every 10 days.

Plotting of cases with two-week projection. Graphic by Peter Khoury.

“They’re staying below that initial exponential rise so that’s good,” Khoury said. “But I don’t want people to be complacent either.”

“A rise is a rise, and as long as it’s going up, it’s not going down,” he continued. “We can flatten the curve but if the curve is still going up, you’re still gonna hit a point of crisis.”

“I know that our political leaders are anxious to get the economy going again, which I think we all are,” Rutherford said. “But it has to be done in a way that’s careful so that we don’t get a big rebound.”

Rutherford brought up what happened in St. Louis during the 1918 Spanish Influenza.

Like San Francisco, St. Louis showed early success in preparing against the surge of the disease. In a couple of months, St. Louis “took their foot off the brake and tried to get everything going again,” he said.

“It was too soon, there was a rebound of deaths,” Rutherford said. “Second curve, that’s what we’re trying to avoid.”

He said that the first thing that people need to realize is that “it’s not gonna go back to normal.”

Similar to how more and more restrictions were put in place in California to get to how shelter-in-place is now, Rutherford said that lifting restrictions would also take a slow and step-by-step approach to avoid a rebound of cases.

“This is not the time to let up,” Rutherford said. “We have done really well so far and we need to keep doing what we’ve been doing so successfully, which is sheltering in place,” he said.

“Hang in there,” Jewell reminds people. “Social distancing is really saving people’s lives.”

Follow our reporting on the COVID-19 numbers in San Francisco here. 

San Francisco’s new data tracker includes data on confirmed cases, testing, and hospitalization throughout San Francisco.

Keep our coverage coming and be well.

Join the Conversation


  1. “We’re not necessarily testing everybody with symptoms,” said Rajiv Bhatia, Assistant Clinical Professor of Medicine at the Stanford School of Medicine.

    Why the F–k not?

    Stanford told us a month ago they would have a test readily and widely available by “Friday”. What’s the hold up?

  2. “…In a couple of months, St. Louis “took their foot off the brake and tried to get everything going again,””

    News today’s – WuHan has let up. We can watch and see what happens. But if history is our guide, we needs be in lockdown for probably the rest of the year, minimum; and maybe another year until a vaccine is developed and distributed.

    That is the rational approach – but also highly unlike.

    Thus, get ready for Round II, Round III, RLR

  3. Locking down for the rest of the year is not rational. Poverty and unemployment leads to lots of death of despair.

    In a couple of weeks we can look at a gradual reduction in lock down, similar to what European countries like Norway are going to do.

    No large crowds, require masks to go into a store. Consider the temperature checks like South Korea does before they allow you in.
    I’ll Trust my opinion over yours as I’ve been wearing a mask and gloves for almost 2 months.

    1. Well, I certainly hope you’re right; and it is true a prolonged lock-down would have devastating Econ consequences for a significant # of ppl – heck, everyone!

      And, as I said, we shall see, re: WuHan lockdown loosening. Apparently Sweden adopted another policy of trying to quarantine vulnerable pop’s instead, but has reversed that (due to pressure more than factual failure, IIUI).

      At any rate, about all we really know is that we have an unprecedented situation, with many unknowns. And after reading that 6′ may not be adequate at all (10-20′ recommended), I doubt we will get to mass innoculation (thus 2nd and 3rd waves) before we see a vaccine widely distributed, worldwide (late ’21-early ’22?)

  4. It’s important to note that the demographics in San Fransisco are skewed towards younger people. If you compare the at risk groups between CA and SF, CA as a whole has 3 times more old people than SF.

  5. The density of NY is much higher than SH, LA and SJ. Comparing Ventilators used 7 in SF and 700 based solely on shelter-in-place is misleading. Comparing LA, San Jose and San Diego is more appropriate. There may be other minor factors such as as temperature, rain, snow should be considered if can be. The experts assessments are major factors but not the only factors.

  6. The government in SF will do everything to justify its actions and claim victory. We can’t let them escape the fact they destroyed hundreds of billions if not trillions of dollars with their actions. For example, AirBnb, a company started in SF (!!!) has been severely damaged by short sighted policies.

  7. I’m sorry to hear about your stock portfolio, Jorge. Airbnb is profitable when they’re not dumping money into their tech stack or marketing, so I think they’re going to be fine. The rest of us are simply glad to not be dying. Thanks for your concern 🙂

  8. Reasons for the San Francisco Bay Area having low Covid-19 infection rates —
    (1) Their primary transportation are personal vehicles; they can afford it. Other U.S. metropolitan areas (outside California) rely on mass transportation, due to lack of income and lack of parking. Cities have reduced subway/train carriages, buses, & schedule frequency in response to social-distancing recommendations.
    (2) Software is their biggest industry, which allows them to work at home. Other U.S. metropolitan areas have mostly service jobs which require workers to go to a place occupied by other people.

Leave a comment

Your email address will not be published. Required fields are marked *