It's hilarious that a high-end New York City restaurant going plant based has triggered so many right wingers. It's almost like they don't trust the market anymore.
As someone who appreciates a well-regulated and vibrant market for food-away-from-home, I'm looking forward to spending some of my forced savings over the past year at Eleven Madison Park next time I'm in NYC!
Cc: @vegan might want to tweet about it ;-)
Technological advances in plant-based food have been quite amazing TBH.

So getting even a few high profile Michelin-rated restaurants to do what Eleven Madison Park did can help influence attitudes and behavior.

• • •

Missing some Tweet in this thread? You can try to force a refresh

Keep Current with Arindrajit Dube

Arindrajit Dube Profile picture

Stay in touch and get notified when new unrolls are available from this author!

Read all threads

This Thread may be Removed Anytime!


Twitter may remove this content at anytime! Save it as PDF for later use!

Try unrolling a thread yourself!

how to unroll video
  1. Follow @ThreadReaderApp to mention us!

  2. From a Twitter thread mention us with a keyword "unroll"
@threadreaderapp unroll

Practice here first or read more on our help page!

More from @arindube

4 May
One possible reason why there are complaints about "labor shortage" is that adding 1M workers when U-rate is 6% has no precedence in pre-pandemic period 1990-2019.

Importantly this is about *changes* ... and doesn't mean that we've reached a high *level* of lab mkt tightness. We're likely to get to much lower u-rate in coming months. But the scale of *change* can explain growing pains and complaints about shortages.

From this perspective:

Temporary Shortage = f(Lab Market Tightness, Rate of Change in Employment)

So temp shortages (even if they were real) may not necessarily imply we've reached maximal tightness, if it's happening at a time of unprecedented rate of change in emp.
Read 5 tweets
24 Apr
IIRC the zero K tax result of Chamley Judd has (at least) two really important limitations.

1) It actually suggests potentially confiscatory tax for a long time, and then eventually reduce rate to zero

2) It assumes infinitely elastic supply of K w.r.t. net-of-tax rate
For the first point, see Straub @IvanWerning:…
For the second point, Piketty and Saez:
Read 4 tweets
11 Apr
Is the $300/week UI supplement holding employment growth back?

Tl'dr: whatever may be happening with vacancies or applications, extending my prior analysis to end of March finds little difference in job growth using x-sectional variation in replacement.
Much of the popular discussion on this seems to have centered around restaurants, which TBH has other things going on: which means "shortages" in restaurants is hardly just about overall labor supply

Here is a thread about some of those possibilities.

However, as @p_ganong points out in this thread, there is an indication that vacancies have risen while job search (proxied by google searches) has fallen - this is a broader point than just restaurants (though would be useful to see broken down).

Read 8 tweets
2 Apr
Rcall makes it fairly straightforward to use packages in R for Stata users. I just wanted to demonstrate with an example of the new Augmented Synthetic Control approach of Ben-Michael, Feller, and
Here's the bit of Stata code to run it.
If you're a Stata user, you should download R, install some of the key packages, and get rcall. It'll make your life easier.
Also, here is a useful set of comparisons of R and Stata commands for Stata users.…
Read 5 tweets
29 Mar
A proof of concept for how modifying Jorda "local projection" w/ clean controls (few lines of code) can produce event studies devoid of negative weighting problems

forval k=0/5 {
g dify`k' = F`k'.y - L.y
reg dify`k' treat i.period if ((D.treat==1) | (treat==0 & Fk.treat==0))
This modifies LP approach to *only* use controls that are still untreated by event date k. Similar to Sun Abraham. But LP approach also readily allows controlling for pre-treatment outcome dynamics and covariates. For example, see Acemoglu et al.
Read 9 tweets
27 Mar
Seems to me another simple way of doing event study w/ staggered DiD is using the Jorda local projections approach modified to admit obs that are either untreated, or first period of treatment.

Something like:
reg (Fk.y-L.y) D i.period if ((D==1) | (D==0 & Fk.D==0))

With this estimator, and example below, I think state 1 will be in the sample (as control) in years 1, 2, and (as treatment) in year 3. State 2 will be in the sample (as control) in year1, and (as treatment) in year2. State 3 will be in the sample (as control) all years.
Inclusion of year fixed effects means we're basically doing a pooled cross section estimation.

Seems to me this provides a consistent event study estimate under HTE with staggered treatment.

But wold like to hear from @borusyak @pedrohcgs @agoodmanbacon and others.
Read 13 tweets

Did Thread Reader help you today?

Support us! We are indie developers!

This site is made by just two indie developers on a laptop doing marketing, support and development! Read more about the story.

Become a Premium Member ($3/month or $30/year) and get exclusive features!

Become Premium

Too expensive? Make a small donation by buying us coffee ($5) or help with server cost ($10)

Donate via Paypal Become our Patreon

Thank you for your support!

Follow Us on Twitter!