Thread by @asymmetricinfo

So let me answer some questions about my medical bankruptcy column: there was so much stuff that just wouldn't fit in! washingtonpost.com/blogs/post-par…

Q) You say this new study excludes someone who "loses their job and has a heart attack from the stress". But those people should be included!

A) No, no, they are included! The point is, you don't want to falsely attribute a bankruptcy to a heart attack. (1/2)

Let's say someone's business failed. Their life's a financial wreck. They consult a bankruptcy attorney, who says "It's time to file." They start getting their paperwork together. The stress gives them a heart attack. The attorney says, "Welp, let's put those bills in the filing"

The heart attack didn't *cause* the bankruptcy, which was already going to happen. But if you just look for medical bills in the filings (the Warren et al method), you'll code this as a medical bankruptcy rather than a business loss.

So instead the authors look at the event: the hospitalization. They look at what happens to you before and after, and ask "How much more likely does a hospitalization make you to declare bankruptcy?" Which is the question we're actually interested in, the *causal* relationship.

That's why their method is better. And also why they get a lower number.

Question 2) Why did you just show data from 2008-2017? We want to look at bankruptcies *before* the recession!

A) Good question! I originally did. But that required me to add a 500 word section on why the data from immediately before the Great Recession doesn't tell us much.

The problem is that a new bankruptcy law took effect in 2005, which made it harder to file bankruptcy. Filings spiked to record levels (2mm+) right before it took effect, as everyone who thought they might even want to file bankruptcy rushed to do so.

Then BK filings fall off a cliff, because almost everyone who wanted to file had already done so. I mean: 2 million in 2005, 600k in 2006 (normal previous level had been around 1.3-1.6 million annual filings.) So they spend 2006 and 2007 recovering from artificially low levels.

Just as we were heading towards whatever the natural level of filings post BAPCPA was, wham! Great recession.

So anyway, the reason I didn't include it was that the 2006-7 numbers aren't very useful. They're artificially depressed, though we don't know how much.

And including them, with little dots to highlight the important dates, just made the graph a total mess. So I cut it out.

Because the important thing is that if half of all bankruptcies were caused by medical bills (i.e. wouldn't happen without the medical bills) ...

Then there should have been a *massive* downward inflection in the number of filings somewhere between 2014-2017, as all of those bankruptcies fall out of sample. There is no inflection at all, just a smooth recovery from the Great Recession peak.

From this we can infer a couple of things:

1) Medical bills are unlikely to have caused anything like half of all bankruptcies.

2) I owe Todd Zywicki an apology.

(Who is Todd Zywicki, and why do I owe him an apology? He was one of the academics whose work--suggesting a high level of "strategic" filings--supported the 2005 reform. I was really skeptical, and thought BK rates would recover to near pre-2005 levels. I was wrong.)

Question 3) How is it morally acceptable that *anyone* be bankrupted by medical expenses! This couldn't happen in Europe!

Answer: Well, no it couldn't, but not for the reasons you think. Contrary to popular belief, people in Europe get medical bills too!

Very few systems cover 100% of all expenses. This is a weird fantasy based on the time someone broke their leg on vacation and totally didn't get a bill! People in Europe have copays, and not every treatment they want is necessarily covered.

If you are on the edge of financial disaster in Europe, a medical bill absolutely could tip you into bankruptcy. Except for one thing: in Europe, it's really effing hard to declare bankruptcy.

It varies by country, of course, from Scandinavians, who are *seriously* draconian about paying your debts, to more relaxed places. But there is no country in the world where it's as easy to file for BK protection as the US. When it comes to BK, we're the bleeding heart liberals.

So you're unlikely to be driven into bankruptcy by medical bills in Europe for the simple reason that most Europeans do not see bankruptcy as a shelter unless you're basically completely destitute.

And with good reason: there is no such thing as Chapter 7 (where you're given a simple discharge of all your debts). Everyone else, everywhere else, has to go on a payment plan. And they're pretty stingy about what they consider basic living expenses, again unlike the US.

But can medical bills add to the distress of a European who is already financially distressed? Yes, just like here. And if Europe had our BK rates, and our easy bankruptcy, there would be some people driven into it by medical bills.

Question 4) How can we not have loads of bankruptcies caused by medical bills? Have you seen how big medical bills can get?

Answer: Good question! Yes, medical bills can get huge. But that's a somewhat deceptive number, because almost no one pays those bills as written.

Your insurer doesn't pay the number on the sheet; they've got some separate negotiated rate. And how many uninsured people do you think pay a $150k bill for a serious surgery? If they had that kind of money, they'd have insurance.

But this brings us back to that sweet, easy US bankruptcy I was just talking about. (to be clear, bankruptcy is not sweet or easy. I'm just speaking in comparison to Europe.)

What does the hospital get if you declare bankruptcy? Nothing, is what they get. Your 401(k) is shielded from creditors (Aside: this is why you *never* break into the 401(k) to pay off debt). A portion of your home equity is shielded. So are various other things.

Most people don't have any significant assets beyond a smidgen of home equity and an underfunded retirement account. If you declare bankruptcy, the hospital will be splitting the $237.85 in your checking account with all your other creditors.

Or maybe they'll be splitting whatever paltry income you're earning with all your other creditors, if you file Chapter 13 instead of Chapter 7. Either way, they know they're never getting that money.

So the hospital actually has a strong incentive to keep you out of BK.

Realistically, you call them up and say "Yeah, I'm a roofer, I make 30k a year, this bill is 5x my annual income, you're never getting that money, how about we work out a payment plan for something reasonable", you have a very good chance of getting a deal.

Partly from financial self interest, and partly because hospitals know that "Squeezes roofers for bill 5x their annual income" is not a good look for them on the local news.

Credit card companies are different. But hospitals are constrained.

Question 5) But medical bills show up in filings!

Answer: Sure. First of all, there are dumb providers who try to squeeze money out of patients who don't have it. But more importantly, when you file BK, what's the first thing you do? Include all your bills!

A bill owed a healthcare provider is an unsecured debt. It quite properly goes into your filing. But that doesn't mean the bill caused the filing.

One example: I did a deep dive on a bankruptcy related to housing debt and overspending. theatlantic.com/business/archi…

There's no question what caused the filing: two expensive divorces, (before this couple married), a house they couldn't afford, and an inability to live within those reduced means.

But there were significant medical bills, IIRC for a cosmetic dermatologist.

Of course they included them! If you're going to BK, you include every single unsecured debt you can remember, no matter how small. Because as @legalinspire points out, if you don't include them, you don't get the BK protection.

And if you're declaring bankruptcy, you're pretty much definitionally struggling financially, so you you want to start your new post-BK life with all your bills wiped out.

But cosmetic dermatology did not cause that BK. You can't infer from presence, or size.

Anyway, this has been another episode of "Megan McArdle explains probably more than you want to know about some reasonably arcane topic". For the shorter version: read the column. washingtonpost.com/blogs/post-par…