Lightnews — Scholar-powered news

2

Vladislav Morozov @vladislavmorozov.bsky.social · Jun 10

My favorite result in this block:

Even with just 2 periods of data, you can identify average causal effects, even if people differ in infinitely many unobserved ways and the outcome function is completely unrestricted.

That's the power of panel data.

Vladislav Morozov @vladislavmorozov.bsky.social · Jun 10

Beyond linearity:
What can we still learn when we don’t restrict functional form and allow arbitrarily rich unobserved heterogeneity?

This new section covers:
• A gentle intro
• Heterogeneity bias
• Average effects via panel data
• Stayers and why they matter
• Local polynomial regression

Why I Switched from Beamer to Quarto Reveal.js for My Presentations

Vladislav Morozov @vladislavmorozov.bsky.social · Jun 10

Just added a new section to my graduate lecture notes — on nonparametric models with unobserved heterogeneity.

It includes one of my favorite identification results in all of econometrics.

1 1 3

Vladislav Morozov @vladislavmorozov.bsky.social · Jun 4

Wrote a short post with details about why it happened and what I like.

vladislav-morozov.github.io/blog/web/qua...

Why I switched from Beamer to Quarto Reveal.js for reproducible, maintainable, and portable slides in teaching, research, and data science.

Vladislav Morozov @vladislavmorozov.bsky.social · Jun 4

I version-control everything with Git, sync and deploy via GitHub, and present directly from a browser.

It’s reproducible, portable, and just works.

Vladislav Morozov @vladislavmorozov.bsky.social · Jun 4

Executable slides: code runs during render, outputs (plots, tables) are embedded automatically.

Simple syntax, responsive HTML, and interactive options too.

7 Variance of Heterogeneous Coefficients – Econometrics with Unobserved Heterogeneity

Vladislav Morozov @vladislavmorozov.bsky.social · Jun 4

I’ve stopped using LaTeX Beamer for slides.

All my research and teaching presentations are now Quarto Reveal.js, and I feel very happy about it.

#EconSky #DataSky

1 3

Vladislav Morozov @vladislavmorozov.bsky.social · Apr 29

Lecture notes here: (new sections are 7-10)

vladislav-morozov.github.io/econometrics...

Or

github.com/vladislav-mo...

Learn how to identify and estimate the variance of heterogeneous coefficients in linear panel models under minimal assumptions (Lecture Notes)

Vladislav Morozov @vladislavmorozov.bsky.social · Apr 29

The main idea: you can identify the full distribution of effects almost as easily as the average!

But these results aren’t widely used — maybe because the original treatment is pretty dense. I tried to make them more accessible via a clean special case.

Vladislav Morozov @vladislavmorozov.bsky.social · Apr 29

Just uploaded the second big chunk of my lecture notes on linear models with heterogeneous coefficients! The notes for this topic are now complete.

This new section goes beyond average effects — to the variance and full distribution of heterogeneous coefficients.

1 3

Vladislav Morozov @vladislavmorozov.bsky.social · Apr 21

Sorry, missed it! Maybe for some very tractable models?

Otherwise, only the usual characterization for misspecified likelihood: that you are estimating the parameter that minimizes the KL-divergence between the true model and the specified one

I usually find it hard to interpret those...

Why Simultaneous (Joint) Tests Instead of Adjusted Multiple Tests?

Vladislav Morozov @vladislavmorozov.bsky.social · Apr 2

Post link: vladislav-morozov.github.io/blog/statist...

Code link: github.com/vladislav-mo...

Why simultaneous hypothesis tests are better — but not always — than adjusted multiple tests

Two-Way Fixed Effects Estimators with Heterogeneous Treatment Effects

2

Vladislav Morozov @vladislavmorozov.bsky.social · Apr 2

Turns out, the answer is only mostly right:

1. Yes, adjusted multiple testing can lead to a huge loss of power.

2. Surprisingly, in some cases, simultaneous testing actually performs worse (though only slightly).

1 2

Vladislav Morozov @vladislavmorozov.bsky.social · Apr 2

Got the same question every time teaching testing in multivariate regression:

"Why a new tests for joint hypotheses? Why not multiple t-tests with adjustment?"

Usual answer: "because power" — always felt vague. I decided to check and wrote a post. (1/3)

#EconSky

1 2

Vladislav Morozov @vladislavmorozov.bsky.social · Mar 21

Very true! Comes down to what you care about.

As an aside, if you drop linearity of the model, OLS — fixed effects models in this case — can give you "bad" weighted averages with potentially negative weights.
Then you really don't have a nice estimand.

www.aeaweb.org/articles?id=...

(September 2020) - Linear regressions with period and group fixed effects are widely used to estimate treatment effects. We show that they estimate weighted sums of the average treatment effects (ATE)...

www.aeaweb.org

2 1

Vladislav Morozov @vladislavmorozov.bsky.social · Mar 18

If a researcher
1. Knows that the effect is non-negative
2. Thinks that the within regression is targeting the ATE,

they will conclude that that there is no effect.

Even if M is very large and there are many people with β_i = M, so you would have a strong effect from intervening on x.

Vladislav Morozov @vladislavmorozov.bsky.social · Mar 18

A simple example: suppose that you have two periods, one covariate x_{it}, and two types for β: some crazy big number M and 0.
1. Units with positive β do not change x.
2. Units with β=0 change x.

The estimand of the within regression is 0, regardless of the proportions of the types and M.

Vladislav Morozov @vladislavmorozov.bsky.social · Mar 18

Good point!

It is a perfectly fine estimand under a linear model — a convex average of individual effects.

The problem is in (economic) practice: people often interpret that as the genuine ATE. Then one may draw wrong conclusions — this effect can have the opposite sign from the ATE.

Econometrics with Unobserved Heterogeneity

Vladislav Morozov @vladislavmorozov.bsky.social · Mar 18

Link: vladislav-morozov.github.io/econometrics...

Learn unobserved heterogeneity-robust causal inference methods — heterogeneous coefficients, nonparametric models, and quantile and distribution regression

Vladislav Morozov @vladislavmorozov.bsky.social · Mar 18

I have learned a lot much from others openly sharing their specialized materials.

It's only fair to offer my epsilon as well and I hope these materials can serve someone.

Econometrica | Econometric Society Journal | Wiley Online Library

Vladislav Morozov @vladislavmorozov.bsky.social · Mar 18

Just uploaded the first block of my lecture notes on econometrics with unobserved heterogeneity! 📊

Introduction and a block on average effects in linear models with heterogeneous coefficients — why standard estimators fail and a robust approach.

Link below.

#econsky

1 1 6

Vladislav Morozov @vladislavmorozov.bsky.social · Feb 18

Example with productivity: onlinelibrary.wiley.com/doi/abs/10.3...

Example with worker skills:
academic.oup.com/restud/artic...

The Jochmans and Weidner paper above cites some more examples.

Firms are more productive, on average, in larger cities. Two main explanations have been offered: firm selection (larger cities toughen competition, allowing only the most productive to survive) and ....

onlinelibrary.wiley.com

Inference on a Distribution from Noisy Draws

Vladislav Morozov @vladislavmorozov.bsky.social · Feb 18

I dunno if it's what you mean, but some other examples are:
1. Firm-level productivity (TFP)
2. Worker skills
3. Teacher value added.

You may care about their distribution, but you have to estimate all these (with noise).

A paper on working with such estimates:
arxiv.org/abs/1803.049...

We consider a situation where the distribution of a random variable is being estimated by the empirical distribution of noisy measurements of that variable. This is common practice in, for example, te...

arxiv.org

Vladislav Morozov @vladislavmorozov.bsky.social · Feb 12

Honestly? Everything just works. It’s fast, integrates with Zotero, and fits my workflow way better.

Still figuring out the best setup, but I'll document it when I find a winning approach.