Hi everyone, I’m here to #TweetABlackEconomistsPaper, and I am very excited to share work by @CarycruzBueno. 1/n https://twitter.com/paulgp/status/1275436970131378176

If you haven’t met @CarycruzBueno, she’s a postdoc at @AnnenbergInst @BrownUniversity and her website is here: http://sites.gsu.edu/cbueno1/ .

Check out her amazing PhD graduation photos on her Twitter! 2/n

I’ll be tweeting about her job market paper, “Bricks and Mortar vs. Computers and Modems: The Impacts of Enrollment in K-12 Virtual Schools on Student Outcomes.”

Ummm, if there ever was a more timely JMP, I cannot think of it. 3/n

The trouble with measuring the impact of virtual school attendance is that families that opt for virtual schooling are self-selecting into virtual schooling, and simple comparisons of outcomes across sector won’t account for that selection. 4/n

. @Carycruzbueno addresses this issue in multiple ways. Her main strategy is to use an individual fixed effects approach, similar to that used in the charter school literature. 5/n https://www.sciencedirect.com/science/article/pii/S0272775711000756

We’ll go into more detail about that, but first, the results. I do not think it will be a shock to the parents of America right now, but virtual schooling is not so hot. Student scores in English/language arts, in math, in science, and in social studies. 6/n

The negative impacts are big, between 0.1 and 0.4 standard deviations (and pretty much the opposite of attending a Boston charter school )!

Fixed effects analysis which compare students from the same sending school within matched cells find similar results. 7/n

A picture is worth a 1,000 words. Test scores dive, especially in math, and stay down, when a student enrolls in a virtual school. 8/n

On the other hand, students who return to brick and mortar campuses, even after two years in a virtual school, almost fully recover their test score losses. 9/n

You might notice in the graphs that there’s a dip in scores in the year prior to entering a virtual school. This isn’t surprising, since students switch schools for a reason. But you might worry that this biases the results and that scores were on a downward trend anyway. 10/n

. @Carycruzbueno addresses this by estimating an interrupted panel (dropping the year prior to entry), again finding large negative impacts. The return to similar scores with the return to a traditional school convincing (2nd graph above). 11/n

While there are still some concerns about identifying off of switchers with the individual FE approach, there aren’t really other good options here! Virtual schools have few capacity constraints (no lotteries) and location doesn’t matter (no distance instruments). 12/n

In short, @Carycruzbueno shows that virtual school is not good for student achievement.

I'm sure parents across the U.S. agree.

The paper is still in the works, so consider this the trailer! If you want to know more, follow @Carycruzbueno & invite her out for a seminar. 13/n

The fun part about doing this with a new paper was that I got to comment on ongoing work. I’d love to include this paper in my syllabus when I teach fixed effects this fall, so I’m hopeful there will be a full public release soon! 14/n

I don’t usually participate in chains, but I was so thrilled to share @CarycruzBueno’s work I couldn't resist.

Here's where I'm supposed to challenge others. I’m hesitant to put folks on the spot because I don’t know what everyone has capacity for right now. 15/n

Instead, I want to challenge you. Yes, YOU, the reader of this tweet. If you’re reading this tweet, it’s pretty likely that you are an economist or economist-adjacent. 16/n

It’s also pretty likely that you committed to yourself, or your students, or your colleagues that you would diversify your syllabus this summer. 17/n

So, since you’re going to be reading papers by BIPOC economists anyway & adding them to your syllabus, get started now! Start by trading out one paper by a white dude and trading in a new paper. And while you’re doing it, #TweetABlackEconomistsPaper.

Yes, you, I mean YOU. 18/18

Thanks @femonomics for getting this started and @paulgp for passing it along.