Feodora Teti
@FeodoraTeti
Followers
608
Following
226
Media
6
Statuses
77
Economist at @ifo_Institut and @LMU_Muenchen | Trade Policy
Munich
Joined October 2021
Don't wanna be here?
Send us removal request.
Explore trending content on Musk Viewer
オーロラ
• 341828 Tweets
Joost
• 296472 Tweets
CHAWARIN ON STAGE
• 179982 Tweets
FOURTH SINGING TIME
• 119373 Tweets
#withMUSIC
• 97434 Tweets
自分これ
• 59310 Tweets
Fulham
• 58418 Tweets
#不破湊3Dライブ
• 49014 Tweets
#العين_يوكوهاما
• 33172 Tweets
太陽フレアのせい
• 30241 Tweets
カクレンジャー
• 27304 Tweets
PERTHCHIMON LIVE x TARO
• 27040 Tweets
Gvardiol
• 26712 Tweets
マリノス
• 24951 Tweets
Buju
• 20838 Tweets
#FULMCI
• 16235 Tweets
自分の魚
• 15328 Tweets
HeavenlyVoice WithMrC
• 15150 Tweets
わたしの誕生魚
• 14385 Tweets
Dremo
• 13867 Tweets
ホールツアー
• 10079 Tweets
Pinned Tweet
Thank you
@jenniferdoleac
@TradeDiversion
and all all those who've shared my work on tariffs over the past days. Your support has been incredible. I'm currently working on a heavily revised version of the draft that will be available later this fall.
3
33
187
Excited to share some personal news: I'll be joining
@IESPrinceton
as a postdoc for the 2024/25 academic year! Can't wait to be part of the vibrant academic community at
@PrincetonEcon
and learn from the brightest minds. Let's connect if you'll be around – looking forward to it!
21
2
229
Gestern hatte ich das Vergnügen unsere
@ifo_Institut
-Studie zu
#lieferketten
, die ich mit
@LisandraFlach
@Steininger_MG
Andreas Baur und Jasmin Gröschl für die
@KASonline
erstellt habe, vorzustellen und mit
@Josef_Braml
& Melanie Vogelbach (
@DIHK_News
) darüber zu diskutieren 1/4
1
7
47
The first week at
@CEP_LSE
@LSEEcon
has been a blast, thanks
@swatdhingraLSE
for making it happen! I will be in London until Christmas, let me know if you are around
3
1
44
However, please understand that a lot of work went into this project and as a junior with a Postdoc contract, I need the accompanying paper to be in good shape and close to submission before I make the data available to everyone.
1
1
38
Dear PhD students: please note, when something looks this off, it probably is. This is how Russian imports of items critical to Russian weapons systems and its military development look like in our work!
Big new report “Back in Stock” from our Russia team on the impact of sanctions on Russia’s defense industry. Bottom line: Russia has adapted with the vital help of Beijing. Note exports in dual use goods from China to Russia skyrocket after Putin and Xi met in March 2023. 1/
7
120
225
2
6
37
I had the honor to spend the past week in beautiful Lindau at the
#LINOecon
@lindaunobel
. Thanks for all the fantastic lectures, the interesting conversations and new connections!
1
1
31
Gute Zusammenfassung über die gestern beschlossenen Sanktionen mit Einschätzungen von u.a.
@kielinstitute
,
@WIFOat
@HO2604
und meiner Wenigkeit
@ifo_Institut
0
8
23
In the paper, I show that tariff data from standard sources are unreliable and the resulting trade elasticities are highly sensitive to the messy data.
1
5
18
Finally, a big thank you goes out to
@WiEMecon
@BeckerFriedman
. It was an amazing conference, I met so many inspiring and great women, and thanks
@m_bombardini
for the brilliant discussion.
0
0
12
My personal highlights: talking to
@metrics52
about a new research idea, discussing soccer with Chris Pissarides and meeting inspiring young economists from all around the world. Big thanks for nominating me
@VfS_econ
@ifo_Institut
@ECONMunich
0
1
10
For everyone who is interested in the data:
I will release a beta version ASAP. Of course, I want to provide the community with better data than we currently have.
1
0
11
But here is the catch: when a preferential tariff isn't reported for a year, it's assumed there is no trade agreement. The result? False interpolation
1
0
9
Why should we care? The gains from trade change dramatically with different trade elasticities, therefore, it matters whether the elasticity equals -5 or -2 or -1. Let me briefly summarize the two main issues of WITS, the standard source for tariff data:
1
1
9
3) I back out the AHS=min(MFN, pref) using the filled tariffs for all country pair-product-year observations to overcome the selection bias resulting in 5.6 billion observations. The solid line shows the result:
1
1
9
Problem
#2
) Selection
Not included in the Feb23 draft!
The second issue cannot be illustrated with pretty graphs, but it is equally problematic. The effectively applied tariff is only available when UN Comtrade, which is also available through WITS, reports positive trade.
1
1
9
In the application, I show that correcting for the missing observations matters: existing estimates of the trade elasticity change drastically. The direction depends on the respective empirical method and sample, more to come in a follow-up tweet as it is still a work-in-progress
1
0
8
How do I fix these issues?
1) I put together all tariff data that I could find to minimize the number of missing preferential tariffs. For roughly half of the agreements in force, I have detailed tariff schedules, so for these, no interpolation is necessary.
1
0
7
Problem
#1
) Measurement error due to false interpolation
Take a look at the graph that shows the simple average of the effectively applied tariff (AHS) that Mexico imposes on the US according to WITS.
1
0
7
By examining disaggregated data by tariff type on WITS, we can differentiate between Mexico's MFN tariff (in blue) and its preferential tariff under NAFTA (in red). The dots in the graph represent yearly averages.
1
0
7
While MFN tariffs are reported almost every year, the reported pref. tariffs have significant gaps. This reveals how the “min-rule” leads to false interpolation when a preferential tariff isn't reported but still exists, AHS=min(MFN, .) = MFN, causing spikes in the time series.
1
0
7
On top of that, we know that also UN Comtrade contains many missing (non-zero) flows (c.f., the great work by CEPII on this). Hence, the sample of available tariffs is “two times selected”, and tariffs of low-income countries will be undersampled.
1
0
7
What is going on? WITS uses the concept of effectively applied tariffs (AHS), defined as the lowest available tariff. If a preferential tariff exists, it will be used, otherwise the AHS equals the MFN.
1
0
6
Especially, when one is not interested in the disaggregated 6-digit tariff data but instead only needs aggregated data (for example ISIC sectors), this leads to a highly selected sample: only tariffs with positive trade flows will be used to average.
1
0
6
Danke an JanCernicky für die tolle Zusammenarbeit! Hier die Studie: 4/4
1
1
6
Tariffs jump between 0-18%, which seems erratic and implausible, especially considering that NAFTA entered into force in 1994.
1
0
6
With these 2 pieces of info and the reported tariffs, I interpolate the missing observations for every country pair-tariff schedule for the entire period 1988-2017. If the agreement allows for phase-in, I linearly interpolate the reported tariffs conditional on having an RTA.
1
0
6
For preferential tariffs, I combine the available tariffs with information on RTAs available through DESTA, Mario Larch, and Bayer & Bergstrand. With this data at hand, I know the entry into force of every agreement and whether that agreement allows for phase-in.
1
0
4
2) I developed an algorithm to fill in the missing tariffs. For MFN tariffs, I set them equal to the nearest preceding observation or succeeding observation if no preceding data is available (same as John Romalis does in his work).
1
0
4
@dix_rafael
I use an Alias for any Starbucks order that I immediately forget and only remember after the barista angrily shouts “juuuuulia” for the 10th time
0
0
4
Stattdessen bietet eine stärkere Diversifizierung der Lieferkette die Möglichkeit länderspezifische Risiken zu minimieren. Aufgabe der Politik ist es die Rahmenbedingungen für Unternehmen zu schaffen, die Diversifizierung erleichtern und so resilientere Lieferketten fördern. 3/4
1
0
4
For agreements without phase-in, I set missing observations equal to the nearest preceding or succeeding observation. This procedure ensures that all years up until the entry into force year are filled up.
1
0
3
@TradeDiversion
thanks for re-tweeting! here is a thread summarizing the paper:
Thank you
@jenniferdoleac
@TradeDiversion
and all all those who've shared my work on tariffs over the past days. Your support has been incredible. I'm currently working on a heavily revised version of the draft that will be available later this fall.
3
33
187
0
0
3
as
@julianhinz
said, I am not claiming that the overall message is wrong, not at all, but please let's do things correctly instead of spreading fake news.
1
0
2
0
0
2
@jenniferdoleac
here is a thread summarizing the paper:
Thank you
@jenniferdoleac
@TradeDiversion
and all all those who've shared my work on tariffs over the past days. Your support has been incredible. I'm currently working on a heavily revised version of the draft that will be available later this fall.
3
33
187
0
0
1
@DanielYiXu
thanks for re-tweeting this! here is a thread summarizing the paper:
Thank you
@jenniferdoleac
@TradeDiversion
and all all those who've shared my work on tariffs over the past days. Your support has been incredible. I'm currently working on a heavily revised version of the draft that will be available later this fall.
3
33
187
0
0
2
Auch wenn Lieferketten aktuell unter Stress sind, ist reshoring nicht die Lösung:
Wir zeigen, dass Risiken durch Unsicherheiten in der Lieferkette zu einem möglichen BIP-Rückgang von fast 10% führen müssten, damit sich eine Rückverlagerung lohnen würde. 2/4
1
0
2
1
0
1
@IngaHeiland
This is great news!🇨🇭is a bit particular, because of the almost exclusive use of specific tariffs, which are a function of prices & can therefore fluctuate a lot. Eliminating these types of tariffs should reduce trade policy uncertainty, too!
0
0
1
0
0
1
@KyleLHandley
So true. Especially because the export bans will only have (if any) an effect in the distant future. As if 🇷🇺 machines used to extract oil will all break down tomorrow…
0
0
1
@KyleLHandley
I think so but my guess is that there is a structural issue with the data that they use. As far as I understand it is not standard trade data such as comtrade but instead they come from a data aggregator such as datamyne.
0
0
1
@CoppetainPU
@DavidKleimann
for Canda, I have found that they will apply a uniform tariff of 35% unless the MFN is higher or goods. At least this is my reading of this notice
1
0
1
0
0
1
@SennJulien
Not all CESifo conferences are open to non-members, no idea how it works in behavioral though
0
0
1
@jameswesleylake
@nd_econ
@rafer77
@YotovG
@jeffbergstrand
@jameswesleylake
your new paper sounds super interesting 👏🏻
1
0
1
@thom_sampson
@hale_shale
I find the different patterns for the ex-/imports particularly surprising because the border customs procedure on the EU-side were in place since day one while the UK slowly phased them in. So I would have expected the pattern to be the other way around
1
0
1