Alright fellow epi and econ friends, gather 'round, time for me to talk about this article

Because I am a native econ whose life quest it is to bridge the epi/econ causal inference methods divide.

And because I recognize a lot of my own errors in it. https://twitter.com/NYTHealth/status/1279674495662465024

To outline:

1) The framing is atrocious, in particular since it implies epi-ists < econ-ists

2) The methods divide is very real, but not for the reasons implied.

3) Epi would, indeed, benefit greatly from embracing these methods, but

4) This article only hurts that effort.

The framing here is really bizarre. It starts with saying that RCTs exist and that other approaches can be used, but ONLY mentions the econ-preferred route.

Epi has an entire field of causal inference with observational data. To not even mention it is negligent.

The implication is pretty clear: either the epi-style approaches are so bad they aren't worth mentioning, or that economists' ideas are innately superior to epidemiologists' when it comes to causal inference.

That's wrong, period.

My epi causal inference colleagues are among the most impressive people I've worked with in any field.

We have very different lenses. I can think my native econ lens has some benefits without assuming the native epi lens is trash (which it absolutely is not).

I recognize myself in this article. In the past, I might have written something similarly framed. It's in part endemic in the way economists are cultured and trained.

I've learned from and continue to purge this kind of erroneous thinking. This article may help you recognize it.

It's absolutely true that epi and econ have WILDLY different sets of methods and approaches and ideas when it comes to causal inference in observational settings.

It's really hard to overstate just how different they are given that both deal with essentially the same problem.

But the reasons they are so different are really complicated, and probably not for the reasons you might think.

Rather than rehash this for the millionth time, I wrote about it a year ago here: https://twitter.com/NoahHaber/status/1133373883560681472

In addition to what's in the thread, epi trainees get little exposure to the quasi-experimental styles of methods.

That is a real issue, and one which I and many others are working to solve.

These methods would be INCREDIBLY useful in epi if they were more available.

The one that stands out the most to me is regression discontinuity, which is based on arbitrary decision thresholds. Those thresholds are EVERYWHERE in medicine.

Epi folks: RDD is a gold mine for kickass studies. Trust me on this one, you won't regret it.

So, I completely and absolutely agree with one of the premises of the NYT article: that epi stands to benefit a LOT from embracing these methods and collaborating with the folks who are most knowledgeable about them.

Let's make that happen.

Of course, no method solves all problems. One of the properties about the econ-style methods is that they only work in very narrow and limited circumstances (see thread above).

The NYT article just kinda...ignores that part.

Causal inference is hard.

But articles like this are absolutely not the way to getting us in epi (I am counting myself in the epi-o-verse) to change.

We economists (I am also counting myself in the econ-o-verse) have to be learning before we can be teaching, or at least do both at the same time.

One of the hardest lessons learned in development econ is that you can't just step into a place you don't understand, without invitation, and "fix" things.

We have to spend the time there. We have to be welcome there. And we have to make ourselves so that we are welcome.

I'll be the first to tell you that I learned a lot of this the hard way, and was only able to survive long enough to learn this and change by the generosity of my friends and colleagues.

A few years ago, I might have written something very similar to this article.

If are looking for something that gets at all the good stuff in this article without all the BS, here is a really nice recent paper from @EMatthay @MariaGlymour and others.

This is how it should be done. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6926350/

One of the article's authors @AnupamBJena's response to @mlipsitch's critical thread (h/t @NeuroStats) https://twitter.com/AnupamBJena/status/1279883173866995715