Megan McArdle

Megan McArdle is a Washington Post columnist and the author of "The Up Side of Down: Why Failing Well Is the Key to Success."
  • Books by Megan McArdle:
  • The Up Side of Down: Why Failing Well Is the Key to Success
  • Buy on Amazon
Recent Articles

MEGAN MCARDLE COLUMN

Advance for release April 15, 2021, and thereafter

(For McArdle clients and FOR PRINT USE ONLY)

For Print Use Only.

By Megan McArdle

WASHINGTON -- On Tuesday morning, the federal government made a big mistake.

The Centers for Disease Control and Prevention and the Food and Drug Administration jointly announced that they are investigating six cases of rare blood clots that may be linked to Johnson & Johnson's coronavirus vaccine - and recommended temporarily halting its use.

Public intellectuals piled onto social media, incredulously asking what the government was doing shutting down a vaccine amid a pandemic over an extremely rare side effect. Longtime FDA watchers, however, saw the agency doing what it almost always does: erring on the side of caution, no matter what.

To be fair to the government, the reported clotting problems are of an unusual type, and heparin, a standard treatment for blood clots, may make these cases worse. Officials are understandably worried that doctors could unintentionally harm their patients.

So one could argue that the government is doing what it should: pausing to assess the extent of the problem and give doctors time to absorb new treatment guidelines. And if a few days' delay were the only side effect, this would be a strong argument.

The Johnson & Johnson vaccine is a small fraction of the U.S. stock and and even that supply was already being temporarilydisrupted by production problems at a Baltimore plant. While it's nice to have more options - particularly ones that don't have the exacting refrigeration requirements of the messenger RNA vaccines from Pfizer and Moderna - this is probably not going to slow the pace of vaccination nationwide much, especially since we're quickly approaching the point where vaccine hesitancy is going to be a bigger problem than shortages.

But that's exactly why the pause is likely to do more harm than good. Europe's experience suggests that when regulators halt vaccinations even temporarily, it sends a negative signal about the vaccine's safety that lingers in the public mind. And even if the clots are a real side effect of Johnson & Johnson's vaccine, in context, this vaccine still appears very safe.

So far we know that six women who got the Johnson & Johnson vaccine experienced a rare form of blood clot, and one died. But more than 7 million people have received this vaccine, which means that if the vaccine is causing this problem, it appears to be a less than one-in-a-million side effect. Extrapolating that out: If everyone in America received this vaccine and it caused clots at the same rate, there would be around 330 such cases, and perhaps some fatalities. By comparison, almost 500 Americans died of covid-19 Monday.

Of course, people make distinctions between passive risks, like catching covid-19, and active risks, such as taking a vaccine. But other common medications and activities can cause blood clots, notably birth control and flying. Most people aren't deterred, because it's a tiny risk against a great gain. That's also true of protecting ourselves and our neighbors from a deadly pandemic.

This announcement does not vindicate the vaccine hesitaters who worry that drug companies and regulators took shortcuts; paradoxically, it's almost a good sign that we're discovering rare side effects.

No pharmaceutical trial is powered to reliably detect one-in-a-million side effects, because that would take millions of trial subjects, which would be less like a trial and more like . . . injecting a substance into millions of people without studying it first. And even if we had possessed data showing such a rare effect, it shouldn't have stopped regulators from approving a vaccine for covid-19, which kills a lot more than one in every million people who catch it.

Based on what we know right now, the math still strongly favors using this vaccine. Unfortunately, the FDA's action has probably made people less likely to take it, or any coronavirus vaccine, even if regulators ultimately pronounce it safe. Even more unfortunately, this is a pattern: When faced with a difficult decision, the FDA almost always chooses the more restrictive option.

The defense of this institutional conservatism is generally that it has prevented the approval of drugs with awful side effects, such as thalidomide. That's true. But economist Alex Tabarrok points out that excess caution has an equally real price, which he calls "the invisible graveyard." It's full of people who died waiting for the FDA to be extra sure about a good treatment.

Only in a global pandemic, when we are all watching the deaths pile up, do we see the grim outline of all those invisible tombstones we've been paying for. And having seen them, we need to do more than just shout at the FDA about this one case; we need to decide right now whether we're willing to pay such costs - and if not, how we're going to stop overcautious regulators from digging more graves.

- - -

Follow Megan McArdle on Twitter, @asymmetricinfo.

MEGAN MCARDLE COLUMN

Advance for release Tuesday, April 13, 2021, and thereafter

(For McArdle clients and FOR PRINT USE ONLY)

For Print Use Only.

By Megan McArdle

WASHINGTON - In the end, it wasn't even close. By late Friday morning, even with roughly a thousand ballots outstanding, so many workers at an Amazon warehouse in Bessemer, Alabama, had voted against joining the Retail, Wholesale and Department Store Union that the "No" tally now represented more than half of all ballots cast.

As journalist and former labor movement strategist Rich Yeselson tartly suggested, unions shouldn't hold elections they're likely to lose this badly. Too, a newly opened warehouse in Alabama might not be the place to start a drive to unionize a national company.

But whatever your opinion about this election -- or unions more generally -- this likely won't be the last big and closely watched campaign to unionize an Amazon facility. (Amazon founder and chief executive Jeff Bezos owns The Washington Post.)

For unions, the company represents something they haven't had since the early 20th century: a private-sector target with the kind of scope, and profit margins, to claim some real value for the workers.

The lack of such targets has been a significant reason the union share of the workforce has been in steady decline since it peaked in the mid-1950s, at around 35%. The left tends to blame Ronald Reagan, but by the time he famously broke the Professional Air Traffic Controllers Organization, union membership had already declined to around 20% of the workforce. Nor can Reagan alone explain why unionization rates have declined almost everywhere in the world since 1990.

What has changed almost everywhere simultaneously is that manufacturing is a smaller share of employment in developed countries. In 1950, 3 out of every 10 American jobs were in manufacturing. In 2020, fewer than 1 in 12 jobs were. And compared with the labor-intensive industries that replaced them, manufacturing is fantastically productive.

Consider two giants: Walmart and General Motors. In 2020, General Motors generated about $790,000 in sales, and $41,500 in profit, for each of its 155,000 employees. Walmart had more than four times the annual revenue of General Motors, but needed almost 15 times as many employees to do so. Sales per employee were only $240,000, and the company's profit margin on those sales is also lower, so net income per employee was actually around $6,750.

That's not nothing, of course! But realistically, a union can't claim 100% of profits for workers, which creates a conundrum for would-be organizers. When net income per employee is that low, it's harder for unions to improve worker compensation net of the dues needed to pay for the union -- especially since a retailer whose employees are flung out across thousands of locations is likely to be more expensive for the union to organize, and represent, than even huge manufacturers with dozens of plants.

Unfortunately for American union organizers, U.S. manufacturing has increasingly been automated or outsourced, and what remains is facing fierce competition from companies abroad. Companies can't just jack up prices to compensate for higher labor costs, as their mid-century predecessors could. That's why unionization is increasingly concentrated in government jobs, which face limited competitive pressure, and which can be centrally organized.

That's also why Amazon represents such an enticing target. Amazon's dominance suggests it ought to have some pricing power. It has hundreds of thousands of new workers who could be brought into the union fold. Its warehouse facilities are even bigger than a Walmart Supercenter, so they're more efficient to organize; they're also significantly more productive than traditional retail. Yet since the company's whole business model now depends on being close to every consumer to facilitate quick delivery, it can't just move operations to a friendlier state, much less Vietnam.

But while unions want to organize Amazon, they still face a steep climb. More than half the company's operating income comes from Amazon Web Services. That money can't simply be transferred to the warehouse workers, since business lines being subsidized by more profitable products are vulnerable to cuts.

Yet take the AWS profits away, and Amazon looks a lot closer to Walmart than GM in organizing potential, even if you assume the company could raise prices a bit - and raising prices would be complicated by the fact that an estimated 60% of sales come from the site's third-party sellers. But an equally daunting problem might be the country that GM made.

Before mass automobile ownership, workers tended to cluster conveniently close together. Today's workers might come by car from an hour away and aren't so easy to reach. The very productivity that makes Amazon financially attractive to organize leaves little time for workers to pause and make friends with their co-workers, building social networks unions can leverage.

Those are structural disadvantages the union is apt to face at whichever Amazon facility it targets. So while the name of the town might be different in future organizing drives, the result might be much the same.

- - -

Follow Megan McArdle on Twitter, @asymmetricinfo.

MEGAN MCARDLE COLUMN

Advance for release Thursday, April 8, 2021, and thereafter

(For McArdle clients and FOR PRINT USE ONLY)

For Print Use Only.

By Megan McArdle

WASHINGTON -- Last week, the Centers for Disease Control and Prevention released its provisional death statistics for 2020, and it gave us a shocking figure: While a reported 345,323 people died of covid-19 last year, total deaths jumped by much more than that. In 2020, the United States lost about 3.4 million people, compared with about 2.9 million in 2019, an increase of roughly half a million. Those extra deaths represent about 1 in every 650 Americans, or, alternatively, the entire population of Atlanta.

It's hard to grasp what numbers like that mean; the imagination trails off into strings of zeroes. But even if you dig beneath the headline data and try to focus on the smaller numbers, you'll find them equally elusive. What they keep driving home is how little we still know about this virus, and what it's done to us -- indeed, how much we may never know.

Consider the 158,000 extra deaths that weren't directly attributed to covid-19. It's safe enough to link most of them to the pandemic, but if you want to know more than that, you quickly run into trouble.

Some of them, of course, are probably just covid-19 deaths that didn't get recorded that way. It seems significant that deaths from strokes and heart disease rose during the pandemic, while cancer deaths held steady. Strokes and heart attacks are known complications of covid-19; cancer is not.

We can probably also blame the virus for 13,000 extra deaths from diabetes, a known covid-19 risk factor. But what about the reported rise in traffic fatalities, most of which probably weren't feverish patients crashing their cars on the way to the emergency room? A more likely culprit is the faster speeds and reckless driving enabled by suddenly open roads.

In some cases, we may never know whether to blame the virus, or our reaction to it. Notably, Alzheimer's deaths increased 10% from 2019. Dementia patients are, of course, especially likely to live in nursing homes, among the places hardest hit by covid-19. But when those nursing homes isolated patients in their rooms to keep the virus from spreading, it was particularly catastrophic for those suffering with dementia. For frail individuals who can't pass their days Facetiming friends or browsing the Internet, isolation was a profound torture that might have sent them into decline even if they never got covid-19.

Moving beyond the medical, we will have to reckon with the sharp spike in homicide across dozens of American cities. How much should we attribute to covid-19, how much to prior trends, how much to the deterioration of police-community relations or some other factor?

Finally, we come to the most surprising category of all: things that actually got better. Deaths from chronic lower respiratory disease actually fell about 3% in 2020, presumably because people with serious respiratory disease were serious about staying home, wearing masks and washing their hands. And despite the harrowing predictions I kept hearing about how lockdowns would trigger a wave of suicides, suicide actually fell by about 6%.

This was so counter to expectations that I reached out to Mark Olfson, a professor of psychiatry and epidemiology at Columbia University. He indicated that many experts also found the numbers surprising. Suicides often follow the kinds of traumatic events that happened a lot during the pandemic: losing a job, a business, a loved one. There also had been a rise in unintentional overdoses and emergency room visits for psychological distress.

But we could name some countervailing factors. Extra-generous extended unemployment benefits, for example, meant that many people who lost jobs were actually better off financially in 2020. In fact, with fewer leisure activities to spend money on, a lot of household balance sheets look very healthy right now, though, of course, some individuals have suffered intensely, especially small business owners.

Meanwhile, some people who ended up living with their families might have had more psychological support than they normally would, and the sense of collective misery might, Olfson said, "buffer it for some people." If you're at home by yourself while everyone is out socializing, you feel like a failure; if everyone is at home binge-watching Netflix, you're one of a crowd.

But Olfson also pointed out that even if the numbers went down in aggregate, there quite possibly were individuals who were driven to suicide by the pressures of the pandemic. The people who looked at struggling friends and family and concluded that we were facing an unprecedented psychological tsunami weren't wrong, exactly; it's just that their experiences weren't representative of the whole country.

For that matter, it's possible that the trauma of this year might have lingering consequences that will show up in mortality reports five or 20 years from now -- the psychological wreckage continuing to wash up well after the storm. We will probably never know quite how bad the hurricane got, only that a terrible thing swept over us, and that much of what it destroyed can never be recovered.

- - -

Follow Megan McArdle on Twitter, @asymmetricinfo.

About
Megan McArdle is a Washington Post columnist and the author of "The Up Side of Down: Why Failing Well Is the Key to Success."
Books
  • The Up Side of Down: Why Failing Well Is the Key to Success
Links