iiinova — where your digital self keeps evolving

Pedro Picchetti @pedropicchetti.bsky.social · 31/10/2025

Hi everyone! I am happy to share this note titled "Difference-in-Differences with Binarized Treatments: A Cautionary Tale". I study the common practice in DiD settings of recoding a multi-valued treatment into a binary indicator. I hope you enjoy it! www.dropbox.com/scl/fi/491xi...

000

Reposted by Pedro Picchetti

VoxDev @voxdev.bsky.social · 17/04/2025

🆕 How regulating junkyards reduced car theft in Brazil Today on VoxDev, André Mancha (@j-pal.bsky.social) discusses how increased regulation of the second-hand automotive part market reduced car theft in Brazil: voxdev.org/topic/instit...

voxdev.org

How regulating junkyards reduced car theft in Brazil

Unregulated markets for second-hand goods fuel crime worldwide. In Brazil, strengthening oversight of the second-hand market for automotive parts helped reduce car theft, simultaneously lowering insur...

03712

Pedro Picchetti @pedropicchetti.bsky.social · 05/02/2025

Exciting personal news!! My Job Market season is over. I am thrilled to announce I am joining PUC-Chile as an Assistant Professor next semester. Looking forward to joining this great department and working at the Católica. Great times are coming🤠

1160

Pedro Picchetti @pedropicchetti.bsky.social · 02/10/2024

Hi all. I'm Pedro Picchetti from INSPER Institute of Research and Education, in Brazil. My #EconJMP provides a new method for the identification of dynamic causal effects using Instrumental Variables. sites.google.com/view/pedropi...

The instrumental variables (IV) method has been widely studied in cross-sectional settings. However, many practical applications involve panel data, in settings where a unit's treatment status may turn on or off over time. I show that in the presence of dynamic treatment effects, that is, if past treatments affect current potential outcomes, standard methods are no longer valid if the instruments are serially correlated. This paper proposes the nonparametric identification of dynamic causal effects in a potential outcomes framework in which potential outcomes depend on the treatment path taken by a unit through time but the first stage is static, in the sense that at each period the IV only instruments its contemporary treatment. I provide a nonparametric estimator that is unbiased over the randomization distribution and derive its finite population limiting distribution as the sample size increase.

04714

Reposted by Pedro Picchetti

apoorva lal @handle.invalid · 30/09/2024

prompted by Jeff Wooldridge extolling the virtues of regression adjustment with propensity weights, implemented it in python and ran some simulations ; it does indeed work pretty well. apoorvalal.github.io/LalQuartoBlo...

apoorvalal.github.io

Comparing selection on observables estimators – Lalgorithms

23210

Pedro Picchetti @pedropicchetti.bsky.social · 24/09/2024

New draft of my JMP coming soon! Crazy roller coaster of emotions so far (about to get crazier), but man what a learning curve it has been! Ten toes down, till the wheels fall off, let's get this thing!

030

Pedro Picchetti @pedropicchetti.bsky.social · 17/09/2024

Dolorido demais ver seu grupo de melhores amigos se desmantelando por conta de mulecagem, triste demais sentir sua rede apoio se desfazendo...

100

Reposted by Pedro Picchetti

NBER @nber.org · 15/09/2024

Synthesizing and critically reviewing the modern instrumental variables literature that allows for unobserved heterogeneity in treatment effects, from Magne Mogstad and Alexander Torgovitsky www.nber.org/papers/w32927

0135

Pedro Picchetti @pedropicchetti.bsky.social · 13/09/2024

Hoje eu tive o prazer de conhecer a UFPE e os talentosos pesquisadores do PIMES!

030

Pedro Picchetti @pedropicchetti.bsky.social · 11/09/2024

Acontece alguma coisa muito louca no aeroporto que as pessoas passam a preferir esperar de pé do que sentadas

000

Reposted by Pedro Picchetti

NBER @nber.org · 11/09/2024

Model restrictions can introduce bias but improve precision. Instead of testing for bias from potential misspecification, it's nearly optimal to adapt, averaging restricted and unrestricted estimates using data-driven weights, from Armstrong, Kline and Sun www.nber.org/papers/w32906

031

Pedro Picchetti @pedropicchetti.bsky.social · 03/11/2023

Had the chance to present my work in progress "Identification in Endogenous Sequential Treatment Regimes" at Brown University today. Super grateful for the opportunity and the feedback.🤠

000

Reposted by Pedro Picchetti

simon heß @hesss.org · 19/10/2023

got tired of doing academic CVs in TeX and made a simple and easy-to-use (r)markdown template. hope this might be useful to some #econtwitter JMCs. let me know if you have suggestions. github.com/simonheb/mar...

57223

Pedro Picchetti @pedropicchetti.bsky.social · 04/10/2023

And just like Paul Simon once said, "the leaves that are green turn to Brown". PS: what a lovely lovely city is Providence.

000

Pedro Picchetti @pedropicchetti.bsky.social · 01/10/2023

Nothing quite like the high of running monte carlo simulations

110

Pedro Picchetti @pedropicchetti.bsky.social · 27/09/2023

My gess is there is no way Jrue Holiday actually stays in Portland. It is not consistent with the rebuild plans. Brad could unload all the picks and find a suitable substitute for Marcus!! That'd be insane!!

000

Pedro Picchetti @pedropicchetti.bsky.social · 26/09/2023

jfc enough with the Damian Lillard rumours!!

000

Pedro Picchetti @pedropicchetti.bsky.social · 23/09/2023

On my way to Boston to watch the Red Sox! Don't know the first thing about baseball, but I know the words to Sweet Caroline so I guess I'll get around😅

100

Reposted by Pedro Picchetti

Adam L @adam-lg.bsky.social · 21/09/2023

"Recent Developments in Partial Identification," Kline and Tamer (2023). Looks like a really nice overview of the state of the partial identification literature. Polisky 📈📉 doi.org/10.1146/annu...

Identification strategies concern what can be learned about the value of a parameter based on the data and the model assumptions. The literature on partial identification is motivated by the fact that it is not possible to learn the exact value of the parameter for many empirically relevant cases. A typical result in the literature on partial identification is a statement about characterizing the identified set, which summarizes what can be learned about the parameter of interest given the data and model assumptions. For instance, this may mean that the value of the parameter can be learned to be necessarily within some set of values. First, the review surveys the general frameworks that have been developed for conducting a partial identification analysis. Second, the review surveys some of the more recent results on partial identification.

0205

Pedro Picchetti @pedropicchetti.bsky.social · 21/09/2023

If you are insterested in Difference-in-Differences with imperfect compliance, Fuzzy DiD, Instrumented DiD or however you feel like calling it, check out my working paper "Marginal Treatment Effects in Difference-in-Differences", with Cristine Pinto papers.ssrn.com/sol3/papers....

papers.ssrn.com

Marginal Treatment Effects in Difference-in-Differences

Difference-in-Differences (DiD) is a popular method used to evaluate the effect of a treatment. In its most simple version a control group remains untreated at

010

Pedro Picchetti @pedropicchetti.bsky.social · 21/09/2023

Hi, Bluesky! My name is Pedro. I am a brazilian PhD student in economics and an aspiring econometrician. My research interests lie in causal inference and semiparametric econometrics. I am also a big sports fan and a fan of hip-hop. If any of these interest you as well, let's link up! 🤠

120