Can Credit Card Spending Predict Coronavirus Spread?

Filed Under: Credit Cards

Credit card companies have released some interesting data about how consumer spending patterns have changed in recent months. For example, in mid-May Chase revealed that credit card spending was down 40%, and that spending in non-essential categories was down 50%.

Could credit card spending actually predict where the next coronavirus spike will happen? A new study suggests so…

Restaurant spending tied to coronavirus spread

A new JPMorgan study suggests that increased spending at restaurants in a particular state is closely correlated to a spike in coronavirus cases about three weeks later. In-person restaurant spending was found to be particularly predictive.

This is according to a study that analyzed 30 million credit and debit card transactions, comparing it to the patterns for new coronavirus cases in a particular state.

According to the research:

  • Arizona, Louisiana, and West Virginia showed the smallest relative decline in restaurant spending compared to the same period the previous year
  • The District of Columbia and Massachusetts saw the biggest decline in restaurant spending compared to the same period the previous year
  • For the week that ended on June 14, transactions were still increasing at full-service chain restaurants in Arizona, California, and Florida, and those states are all seeing big increases in cases

Not surprisingly, the National Restaurant Association has said that while it’s concerned about the increase in coronavirus cases, it’s “irresponsible” to pin the rise in cases on restaurants:

“It is irresponsible to pin the rise on a single industry. Restaurants have historically operated with highly regulated safety protocols based on the FDA’s Food Code and now have taken new steps to meet social distancing guidelines required by state and federal officials. We all have responsibility for wearing masks, washing hands, and social distancing.”


An increase in restaurant spending correlates to an increase in coronavirus cases

Grocery story spending tied to coronavirus spread slowing

Higher spending at grocery stores has been found in states where we’ve seen a slower spread of coronavirus. While most states are seeing an increase in grocery store spending, those seeing the biggest increases have largely seen the biggest decrease in new cases. As of three weeks ago:

  • Grocery store spending was up 20% or more from the same period last year in New York and New Jersey
  • Grocery story spending was up less than 10% in Arizona and Texas

An increase in grocery store spending correlates to a decrease in coronavirus cases

Correlation does not imply causation

While I find this study to be interesting, I think it’s important to point out that correlation does not imply causation. No one is suggesting that restaurants as such are responsible for a majority of the spread of coronavirus.

Rather the general reopening of economies, people gathering more, people being more social, etc., is correlated to the spread of coronavirus. While we don’t dine out at a lot nowadays (thanks largely to Ford’s newfound cooking skills), personally I feel comfortable dining out if I can sit outside and all proper protocols are being followed, while I’m less comfortable indoors.

I don’t think there’s anything about this research that comes as a surprise, though it’s still interesting to see it presented so clearly. It’s also no surprise that restaurant spending has increased in the past few weeks, given that in many areas restaurants only started opening again for dine-in customers in recent weeks.

It’ll be really telling to see how if this correlation continues in the coming weeks.

Bottom line

Increased spending at restaurants is happening in states where coronavirus cases are increasing, while the biggest increase in spending at grocery stores is happening in states where coronavirus cases are decreasing. That’s not too surprising, and seems to be more reflective of whether people are staying home or being social.

Hopefully more of this data is released in the coming weeks, as I imagine we’ll see similar trends.

Does this data on the correlation between spending and coronavirus surprise you?

Comments
  1. Correlation does not imply causation should be on the top of this semi-click bait post.

    It is widely suggested the lack of pirates cause global warming. (not Trump, lol)
    There is even evidence to support that !
    Somalia is among countries with most pirates in the world.
    Somalia is among countries that produce the lowest carbon emissions in the world.

    Does data on the correlation between decreasing pirates and global warming surprise you?

  2. A very credible hypothesis is that people who spend a lot on restaurants, an activity that exposes them to higher risk of contagion, do take less precautions overall and therefore help spreading the virus in restaurants and outside of them.

    And yes, enclosed spaces (which include many, but not all, restaurants) are tied to much higher risks of contagion. Plenty of studies around that, including from the Diamond Princess, all confirming such risk.

  3. A more interesting follow-up would be an analysis of restaurant spending in Blue states vs. Red states…we all know what that data would suggest.

  4. @Mark – bring it on – why do I deserve banning? Let’s see…I’m not racist…I’m not denying the pandemic…I don’t declare wearing masks infringes on my personal freedoms as a ‘Murican…let’s hear it.

  5. Restaurants in BC reopened over 5 weeks ago, cases numbers are flat lining at this point, no spike. Pretty much blows a hole in this “study”.

  6. @UA NYC
    By dividing the US between blue vs red you are actively supporting dividing this country. Good thing we have our president getting another 4 years.

  7. @majik – They said the same thing here in Texas 5 weeks after restaurants opened. “Where’s the rise in cases? There’s no rise!”

    And now, three weeks later, we’re out of ICU space in Houston, hospitalizations statewide have nearly quadrupled, and things are just getting worse.

    Don’t speak too soon.

  8. @KD Gee if only there were some massive gathering of 50,000 people that might have happened in Houston a few weeks ago that could explain why there’s a sudden surge in cases…

  9. I don’t know why any of this comes as a surprise. Everyone should have known from the beginning that a reopening of the economy and — in particular, business that promote social interaction — was going to lead to a dramatic increase in COVID-19 cases. This serves as a friendly reminder that all steps taken thus far have been to slow the spread, not stop the spread. I hope no one ever expected that this virus was going to simply disappear. If you haven’t gotten it yet…it’s waiting for you.

    It would, however, be interesting to further analyze the correlation between this spending and the local policies in place for mask wearing. Where and under what conditions are masks required in locations now opening back up? If they’re required in all public settings, but COVID-19 cases are still on the rise, this could offer some useful empirical insights into the practical effectiveness of wearing masks.

  10. I would certainly buy the argument that restaurant spending is an instrument for the direct causal factors (that is, social gatherings, especially without masks). I suspect that a better instrument would be spending on alcohol consumed on premises, but depending on the spending categories available to Chase, the data may be too messy.

  11. Matthew, regarding “Gee if only there were some massive gathering of 50,000 people that might have happened in Houston a few weeks ago that could explain why there’s a sudden surge in cases…” Since primarily young folk attend BLM protests, should we expect to see a rise in COVID-19 cases among young folk in cities with BLM protests? Since blacks are a high percentage of BLM protest attendees, and at high risk for COVID-19, should we expect to see a rise in cases among those young+black, especially young+black in higher percentage black population cities with BLM protests? Since elderly blacks (especially female) have higher risk factors (asthma, etc.) should we expect to see a second wave rise in COVID-19 deaths among those elderly+black especially elderly+black+female, especially in higher percentage black population cities with BLM protests? If so, will the press finally acknowledge that encouraging BLM protests during a pandemic endangered and unnecessarily killed many members of the black population? Perhaps the mayors of New York and DC should be painting tombstones on their streets instead of kudos to BLM.

  12. @AlohaDaveKennedy

    You should learn how to use space and paragraphs, it very hard to read.

    Regardless of BLM movement (lol, you use protest, what a hypocrite), it is not PC to blame BLM hence you will never see politicians blame BLM for COVID spread. But I do encourage everyone to look at the stats and make their own conclusions.
    Remember “Correlation does not imply causation” don’t blame the lack of pirates for global warming.

    What people need to realize is that @Roger is spot on.
    “all steps taken thus far have been to slow the spread, not stop the spread.”

    Unfortunately, most people doesn’t get it and expect “that this virus was going to simply disappear”.

    Weak leadership and ineffective shutdown just wasted the country a whole 3 month and we end up, a little better prepared but still the same situation as in March.
    Look towards Asia for examples of a more effective control.

  13. @Mark – might want to check in the mirror and the red MAGA hat you are likely wearing…that is the definition of dividing.

    Your boy is done…things were fine and dandy until an actual crisis happened. Biden will bring in his team of adults thankfully to clean up the mess.

Leave a Reply

If you'd like to participate in the discussion, please adhere to our commenting guidelines. Your email address will not be published. Required fields are marked *