col-xs-12
col-sm-12
col-md-12
col-lg-12
col-xs-12
col-sm-12
col-md-12
col-lg-12
videoType
dynamic-media
videoDmUrl
https://delivery-p136806-e1377785.adobeaemcloud.com/adobe/assets/urn:aaid:aem:5077701a-2e01-4862-94f0-cd99390c7d3d/play?assetname=WDR2025_Navigating_Standards.mp4
keyFrameImage
timestamp

00:00 So welcome everybody.

00:01 This is,

00:02 I think,

00:02 um,

00:03 The 3rd

00:05 of our seminar series for the World Development Report

00:08 on Standards for Development.

00:11 And uh today we have uh Camille

00:14 Reberti

00:16 I,

00:16 I pronounce it correctly

00:19 uh Camille is a postdoc.

00:21 She has a postdoc position at the

00:24 at the graduate uh institute in Geneva.

00:27 And she's been working on

00:30 On,

00:30 um,

00:31 standard related issues to,

00:33 to trade,

00:34 um,

00:35 and,

00:36 um,

00:37 we asked her to do a bit of an unconventional presentation

00:41 where

00:42 she will present,

00:43 um,

00:45 uh,

00:45 two papers,

00:46 two different papers,

00:47 so we will have a bit of a break from one to the other,

00:50 but

00:51 very much related to,

00:53 uh,

00:53 the impact of standards around trade.

00:56 The first one is on protection,

00:59 our protectionism,

01:00 the effect of technical regulations on input sourcing.

01:03 And the second one is about adopting voluntary standards.

01:08 Uh,

01:09 so it's two different topics but very much related to,

01:12 to the impacts of of standards

01:14 on trade.

01:15 We let her.

01:17 To present for 45 minutes,

01:19 uh,

01:19 please,

01:19 if you have any questions,

01:21 uh,

01:22 clarification questions first,

01:24 we can,

01:25 you can do it or send it on the,

01:27 put it on the text box,

01:28 uh,

01:29 but otherwise we'll might make sure that we have a good 15 minutes or more

01:33 for Q&A

01:34 at the end,

01:36 OK?

01:37 So thanks Camille for doing this and,

01:39 uh,

01:39 over to you.

01:41 Um,

01:41 thank you very much.

01:42 So,

01:42 good morning,

01:43 everyone.

01:43 Um,

01:43 I'm really happy to be here and present,

01:46 um,

01:47 Today,

01:48 so as um Chavi mentioned,

01:49 today I'm gonna present two different papers.

01:52 So the presentation,

01:53 um,

01:53 it's called Navigating Standards,

01:55 The Impact of Technical Regulations and Voluntary Certifications

01:59 for firms trade um strategies.

02:05 So as I'm sure you're all aware,

02:06 global trade is nowadays increasingly shaped by standards

02:11 such as non-tariff measures and voluntary standards.

02:15 These different measures can enhance product quality and consumer trust,

02:20 but they may also create some barriers to trade,

02:24 despite their either voluntary or non-discriminatory nature.

02:30 So as a result,

02:32 understanding how such rules and regulations affect firms.

02:37 Decisions,

02:38 behavior,

02:38 input sourcing decisions,

02:40 or more trade performance in general is essential.

02:44 And um in this presentation,

02:47 I'm gonna focus on,

02:49 I'm going to present two different papers.

02:50 So the first one is gonna

02:53 be on the effect of

02:55 uh technical biers to trade,

02:57 and the second one is going to focus more on um

03:00 voluntary standards.

03:02 Um.

03:03 So we're a bit

03:05 They are a bit different,

03:07 um,

03:07 as one is mandatory,

03:08 whereas the other one is voluntary,

03:10 but I think,

03:11 um,

03:11 they sort of complement each other in,

03:14 in a way of,

03:15 um,

03:16 understanding more globally how um

03:18 different types of regulation works and affects them.

03:23 Um,

03:24 so I'm gonna start with the first paper that

03:25 looks at the effects of technical virus trade,

03:28 and then I'll move on to the second paper.

03:32 So,

03:32 we,

03:33 we started working on um the first chapter.

03:35 So,

03:36 um,

03:37 bear in mind these figures might change a bit now that uh with the new

03:42 Trump tariff,

03:42 but,

03:43 um,

03:43 at least up to recent years,

03:46 um,

03:46 we observed a large decrease in tariff protection.

03:50 Um,

03:51 all over the world and it's even more

03:54 true when we distinguish between type of products.

03:56 So if you look here at the blue line,

03:58 you can see

03:59 that's even more stringent for um intermediate good where the um,

04:05 The tariff rate is even lower than for

04:08 for the average product.

04:10 On the other hand,

04:11 if we look at a similar picture before technical barriers to trade instead,

04:16 we can see that they have been over the same period,

04:19 they have been growing.

04:21 And again,

04:21 if we look at the blue line,

04:23 um,

04:24 it's even more striking for intermediate goods.

04:27 So here,

04:27 um,

04:28 you have the three different types of products and

04:32 you can see the largest increase are the,

04:35 the products that have been targeted the

04:37 most by technicalvis trade or intermediate goods.

04:43 Um,

04:44 before going a bit,

04:45 um,

04:45 deeper into the presentation,

04:47 just a quick reminder of what,

04:48 um,

04:49 technical barriers to trade is.

04:51 Um,

04:51 so,

04:51 TBTs are all the technical standards

04:54 that differs from international one and might,

04:57 might therefore impact trade.

04:59 Um,

05:00 to give you a more concrete example,

05:02 part of the new Green Deal,

05:04 the EU imposed a TPT on

05:07 batteries

05:08 and accumulators.

05:11 And

05:12 where

05:13 In the regulations,

05:15 if you want to use batteries,

05:17 you have the EU impose requirements on the

05:22 The,

05:24 uh,

05:24 level of hazardous materials that have been using.

05:28 The level of um

05:31 Of recycled materials that I have been using,

05:34 and as you can see in the pictures,

05:35 the batteries are um

05:38 Widely used in different types of products.

05:40 So it could be for electric vehicles,

05:42 it could be for photovoltaic,

05:45 um,

05:45 products,

05:46 etc.

05:46 So,

05:47 um,

05:47 one specific activities might

05:49 have a more global effect on across different,

05:53 um,

05:53 supply chains.

05:56 So,

05:56 what we are interesting in this paper and what we want to look at is,

06:00 um,

06:01 so do TBTs

06:03 cause a diversion of supply chains

06:05 and what could be the drivers um of these GVC disruptions?

06:10 So,

06:11 unlike tariffs,

06:12 standards are non-discriminatory,

06:14 uh,

06:14 but they might have different effects on trade.

06:18 So,

06:18 first,

06:18 um,

06:19 the TBTs might reduce information of

06:22 information asymmetries.

06:23 Um,

06:24 either about the quality of the product or,

06:27 um,

06:28 about whether the product is safe to use.

06:31 And in that case,

06:33 it might thus increase international sourcing.

06:36 However,

06:36 on the other hand,

06:38 um,

06:38 TBTs codify different technology that,

06:41 uh,

06:41 foreign firms might not have access to or might have

06:44 access to other higher costs than dome domestic firms.

06:48 And in that case,

06:49 we would observed a decrease in international sourcing.

06:53 Um,

06:54 so what we wanna do and what I wanna show you is,

06:56 um,

06:57 what could be the effect of a TBT.

07:01 Um,

07:02 so,

07:02 the

07:03 main contribution of this paper are threefold.

07:06 Um,

07:06 so first,

07:06 to answer this question,

07:07 we're gonna create,

07:08 we created a new database that covers all the different,

07:12 um,

07:12 TBTs that have been notified,

07:15 uh,

07:15 between 1995 and 2019.

07:19 We're going to use product trade flows to estimate the effect

07:22 of a TBT on the share of EU import sourcing.

07:26 And

07:27 But,

07:28 um,

07:29 we also include a more vertical mechanisms which I would not go,

07:32 um,

07:33 too

07:34 deep into details,

07:35 um,

07:35 for this plantation,

07:36 but the idea that we have is we're gonna incorporate

07:40 adaptation costs,

07:41 um,

07:41 to the TBT into a model of sourcing.

07:45 Um,

07:46 we're also going to create a new measure,

07:48 um,

07:48 using text analysis techniques.

07:51 Uh,

07:51 we're going to

07:52 offer a new measure of adaptation cost,

07:55 and we're going to rely on French custom data,

07:58 um.

08:00 To show evidence of the microeconomic mechanisms of sourcing decisions.

08:09 Um,

08:10 So to give you an idea of how it works,

08:12 um,

08:13 and of the institutional framework of WTO,

08:16 um,

08:17 sorry,

08:17 of TBT,

08:18 um,

08:18 so at the multilateral,

08:20 multilateral levels,

08:21 um,

08:22 all the WTO members have signed the TBT agreements in 1995,

08:27 and in

08:27 these agreements,

08:28 they agreed to notify to the WTO all the

08:32 technical standards that differs from the international ones.

08:36 At a more regional level,

08:38 um,

08:39 the EU has signed several PTAs with deep integration clauses.

08:45 And in that case we're going to be interested in whether the EU

08:49 has signed trade agreements which includes a clause on either harmonization of TBT

08:56 or mutual recognition of TBTs.

09:00 And even within the EU there are two different regimes,

09:04 so around 70% of the production in the EU is within a harmonized sector,

09:10 where in that case all the standards and all the

09:12 rules and regulations are harmonized across all the EU members.

09:17 But you also have the 30% remaining of the EU

09:21 production that is in the non-harmonized or partly harmonized sector,

09:26 which in that case,

09:27 only mutual recognition applies across the EU members.

09:36 Um,

09:36 so,

09:36 as I mentioned in the introduction,

09:38 um,

09:39 we're gonna use,

09:40 um,

09:40 several

09:41 types of,

09:41 um,

09:42 data.

09:43 The first one is,

09:45 um,

09:45 the new database that we created on all the

09:48 different TBTs that have been notified to the WTO.

09:51 So we collected information on around 30,000,

09:55 um,

09:55 TBTs around,

09:56 uh,

09:57 between 1995 and 2019.

09:59 Um,

10:00 however,

10:00 out of this,

10:01 um,

10:01 30,000 TBTs,

10:04 around 40.

10:05 8% have no um

10:08 product information.

10:10 So,

10:10 we implemented the multi-step

10:12 procedures to recover the product information.

10:14 So either we're gonna use um different sources,

10:17 so we're going to look into the indemned,

10:19 the notifications,

10:20 different official documents.

10:22 Um,

10:23 we're going to match the ICS to the HS codes,

10:26 um,

10:26 and in the last step,

10:27 we're gonna use a machine learning algorithm.

10:30 Um,

10:31 to have,

10:32 uh,

10:32 using the desk text descriptions,

10:35 um,

10:36 to match

10:37 a certain notification to a certain,

10:39 um,

10:40 product information.

10:43 And,

10:43 then on the more regional level,

10:45 we're going to use the World Bank database on provincial trade agreements and um.

10:51 To have information on

10:54 um

10:54 DBT clauses that are included in the trade agreements.

10:58 And

10:59 for within the EU we're gonna use um

11:02 data from the EU Commission that give us

11:04 the list of non-harmonized sectors to distinguish between the

11:08 products that are fully harmonized and the products

11:10 that are not fully harmonized within the EU.

11:14 Um,

11:15 and finally,

11:15 we're gonna use,

11:16 um,

11:17 French firm level data,

11:19 um,

11:20 which allow us to have,

11:22 uh,

11:22 to use a firm level definition of intermediate goods.

11:25 So,

11:25 we're gonna define an intermediate good as a good that is imported,

11:29 um,

11:31 And that is um

11:33 in a different

11:34 HS codes than the core business of the firm.

11:39 Um,

11:39 with the idea of,

11:41 um,

11:42 A different product may be an input for a firm,

11:45 but might,

11:45 might also be,

11:47 um,

11:48 a main export for the firm.

11:49 So,

11:49 we wanted to distinguish between

11:51 a firm that is going to import,

11:53 um,

11:54 specific products for its,

11:56 uh,

11:56 for its production versus,

11:59 um,

11:59 a firm

12:00 that is going to use this product as its main exports.

12:06 Um,

12:08 So,

12:08 if we look at the aggregate level,

12:11 um,

12:12 what we can see is,

12:13 um,

12:14 so if we look at the effects of a TBT imposed by the EU on the share of EU

12:19 that is sourced from within the EU.

12:22 We can see

12:25 activity increases the share of imports that is source within the EU.

12:30 However,

12:31 if we look at the effect for intermediate goods,

12:33 uh,

12:33 we observe that this is only 1/4 of the effects.

12:37 Um,

12:38 so it suggests that sourcing links are

12:40 harder to substitute for intermediate goods.

12:46 Um,

12:47 so just

12:48 one quick slide on the theoretical mechanisms and what we have in mind.

12:52 Um,

12:54 so we decided

12:55 to add,

12:56 um,

12:56 the introduction of a TBT into a source model.

13:00 And here,

13:01 what we have in mind is

13:03 a given TBT

13:04 is going to require um suppliers to adapt their production.

13:07 So if um

13:10 A given country or a given supplier still want to export to the EU.

13:15 They're going to have to adapt that production

13:17 to make sure that they fulfill the new requirements imposed by the EU.

13:21 And um

13:23 we think this adaptation cost is going to

13:26 be country specific and it's going to depend

13:29 on how similar or dissimilar your

13:32 production or your,

13:33 or the regulation in your country or from

13:36 the ones in the EU.

13:38 And um this adaptation cost is going to affect

13:42 the input prices directly through an increase in the unit cost,

13:46 but also indirectly

13:49 through an outside option.

13:51 Um,

13:52 so what we have in mind with this,

13:54 um,

13:55 vertical mechanisms is,

13:57 um,

13:59 So if we assume that um supply chains were previously located in Country A,

14:04 um,

14:05 now we're gonna be interested in whether um

14:08 the country in you,

14:09 in which you imports from.

14:11 It's gonna change um

14:13 following the introduction of a DBT.

14:16 Um,

14:17 so if

14:18 the

14:19 wage in country A,

14:21 um,

14:22 that is now taking into account,

14:24 um,

14:24 it uh,

14:25 so our adaptation costs to the TBT is still lower than the wage,

14:29 um,

14:29 in Country B.

14:31 We,

14:32 um,

14:33 we expect no change in the supplier origin,

14:35 but we're still gonna,

14:36 um,

14:38 um.

14:40 Still think they're gonna be an effect on the prices.

14:44 Um,

14:44 however,

14:45 if now

14:46 the wage,

14:47 uh,

14:47 that includes the adaptation costs

14:49 in Country B becomes smaller,

14:52 um,

14:52 than the one in Country A,

14:54 we now,

14:54 um,

14:55 accept,

14:56 um,

14:56 expect,

14:57 expect,

14:58 uh,

14:58 ships from suppliers from Country A to Country B.

15:02 Um,

15:06 Um,

15:06 so,

15:07 to answer this question,

15:09 Um,

15:10 the research designs we're going to have is,

15:12 so first we're going to look at different firms' import margin.

15:17 So we're going to look at import expenditure and import prices.

15:22 And on the more supply chains reallocation side effects,

15:26 we're also going to look at the effects on

15:28 quantity,

15:29 entry and exit.

15:32 And

15:33 the way we're going to do it is,

15:35 so our first variable of interest is going to be TBT,

15:38 which is going to be equal to one,

15:40 if the EU introduced a new TBT on a given product at a given

15:45 time.

15:47 And then we're gonna look at um

15:50 Whether this effect is different depending on the adaptation cost.

15:54 So it's a

15:56 Um,

15:56 TPT.

15:57 We're gonna measure our EA TPT in two different ways.

16:00 So the first one,

16:01 the first way we're going to do it is to,

16:04 yes.

16:05 I mean,

16:05 sorry,

16:06 sorry,

16:06 can,

16:07 can we ask questions?

16:08 Sorry,

16:08 I,

16:08 I'm,

16:08 I'm not sure,

16:10 yeah,

16:10 yeah,

16:11 OK,

16:11 so,

16:11 so sorry,

16:12 just a quick point here.

16:13 So,

16:14 uh,

16:14 so TDT.

16:16 Um,

16:16 you know,

16:17 basically setting,

16:18 uh,

16:18 a TBT,

16:19 the EU introducing a new TBT is,

16:20 is

16:22 I mean,

16:22 there's a reason why they might do that,

16:23 right?

16:24 So,

16:24 so,

16:25 uh,

16:25 so,

16:26 you know,

16:27 maybe they want to protect,

16:28 maybe they want to,

16:30 uh,

16:30 I don't know,

16:30 maybe there's a,

16:32 uh,

16:32 uh,

16:32 uh,

16:33 you know,

16:33 safety scare and so they,

16:35 they,

16:35 they,

16:36 so,

16:38 So,

16:38 so I guess,

16:38 you know,

16:39 are you dealing with,

16:40 with this potential endogeneity or,

16:42 or are you gonna tell us,

16:44 uh,

16:45 uh,

16:45 you know,

16:45 whether these,

16:46 you know,

16:46 DBTs tend to be introduced in cert at certain times in certain sectors?

16:50 Uh,

16:52 yeah,

16:52 um,

16:53 yeah,

16:53 what,

16:53 what's your,

16:54 what's your take on that?

16:54 I mean,

16:55 so,

16:55 this is like,

16:56 you know,

16:56 they're exhaustionously uh given,

16:58 but,

16:58 but clearly they're not,

17:00 right?

17:00 So,

17:01 yeah,

17:01 so.

17:04 Um,

17:05 so one way we have of dealing with this is,

17:08 um,

17:09 so sometimes when they notify the TPT to the WTO,

17:13 they also give an objective of the measure,

17:15 and we also look at whether the effect is still there

17:19 if we control for

17:21 either if the TPT as an objective of harmonization.

17:26 Of,

17:27 of trade enhancing,

17:29 um,

17:31 Um,

17:31 and there we still have our

17:34 negative effect of the other

17:36 TBTs.

17:37 Um,

17:39 we try to look at,

17:40 um,

17:41 The different effects by different um objectives of the TBT,

17:46 um.

17:48 We also found,

17:49 um,

17:50 so I'm not sure I put it on the slides,

17:51 but even if the objective is protection of the environment or,

17:55 um,

17:56 organic certification,

17:58 something like this,

17:59 um,

17:59 the effect is also sometimes negative.

18:02 Um,

18:03 So,

18:04 the one we looked at were more whether the objective was trade enhancing,

18:08 harmonization,

18:10 um,

18:12 But we also found,

18:13 I think when we looked at either

18:15 uh either protection of the environment or consumer protection,

18:19 I think it was also one objective,

18:21 and there the effect was still

18:23 negative.

18:25 And,

18:25 and sorry,

18:25 just a,

18:26 a related point here.

18:27 So,

18:28 so TPTs now are,

18:29 are,

18:30 are sort of dummy,

18:31 dummy measures,

18:32 but,

18:32 but there could be

18:34 You know,

18:35 there could be,

18:35 uh,

18:36 you know,

18:37 these changes could be quite different,

18:38 right?

18:39 Relative to what,

18:39 uh,

18:40 what the,

18:41 what the standard was,

18:42 uh,

18:43 before the,

18:43 this introduction,

18:44 right?

18:44 I mean,

18:44 so,

18:45 so it could be a big jump,

18:46 jump or it could be a smaller,

18:48 a smaller jump so that,

18:50 so,

18:51 so that,

18:51 you know,

18:51 the,

18:52 the,

18:52 the costs for compliance,

18:55 the compliance caused by the,

18:56 by the exporter,

18:57 uh,

18:57 could be,

18:58 could be quite different.

18:59 So,

19:00 Yeah,

19:00 so,

19:01 uh,

19:01 uh,

19:01 yeah,

19:01 are you,

19:02 are you looking at that or,

19:04 or that's maybe not a relevant margin?

19:06 Yeah,

19:06 what,

19:07 what,

19:07 what,

19:07 what,

19:07 what do you,

19:08 what do you think?

19:09 Um,

19:10 so,

19:10 in the baseline,

19:11 we use a dummy.

19:13 One robustness check that we do is to use account numbers.

19:16 So if a product is targeted by

19:19 10 TBTs that might be different than if it's only 2.

19:22 So that's one thing we do.

19:24 Um,

19:24 and then on the slide afterwards,

19:26 I'll show you how we measure this adaptation cost.

19:29 And the way we do it is to compare,

19:32 um,

19:32 the regulations,

19:33 the newly regulation that is introduced versus,

19:37 um,

19:37 the regulations that is in place in the country,

19:41 um.

19:44 Maybe one additional thing we could do is to compare.

19:49 In between the EU regulations,

19:51 cause I think that's what you meant,

19:53 so,

19:54 um,

19:56 Yeah,

19:57 that's something we haven't looked at.

19:58 So,

19:59 we checked

20:00 every new regulations,

20:01 how different it is from

20:03 what is applied,

20:04 but not

20:05 how different the new one is compared to the other one.

20:10 Um,

20:11 I think we tried to do it.

20:13 So instead of,

20:13 uh,

20:14 comparing the new ones to the one that is in place,

20:16 we also compare

20:18 the,

20:19 all the EU measures,

20:21 including the new ones to everything.

20:23 Um,

20:23 and I think the results were quite similar.

20:31 Or maybe we could look a bit more specific into changes in between the regulations,

20:37 in between duties.

20:46 OK.

20:47 Um,

20:49 So the first way we measure the adaptation cost is to compare um

20:55 the effect the effect that TBT is gonna have um

20:58 within the EU compared to outside of the EU.

21:01 We have the idea of um

21:04 The EU suppliers might have more um incentive to

21:09 Uh,

21:09 to comply with the new regulations because even if

21:12 they only want to serve the domestic market,

21:14 they have to,

21:15 um,

21:16 comply with the new regulations versus

21:18 an extra EU suppliers,

21:20 um,

21:21 depending on how big the EU market represents,

21:24 might not be willing to,

21:26 um,

21:26 implement the new standards.

21:29 And the second way we're going to measure our adaptation cost is to look at

21:36 how dissimilar the

21:39 The regulation in the EU is compared to the

21:42 regulation in the country you want to import from.

21:45 So here to give you an example,

21:46 um,

21:47 we're gonna compare,

21:48 so,

21:49 let's say TBT3 is the new EU TBT that is imposed.

21:53 And uh we want to compare it to

21:56 um

21:57 to how

21:58 distance it is from the irrigation that is in place in Australia.

22:03 So in that case,

22:03 we're gonna compare the contents of the TBT of the EU TBT

22:09 to the content of um

22:11 the TBTs imposed in Australia.

22:15 So,

22:15 here we put below um

22:17 an example.

22:18 Um,

22:19 so TB23 is now um is looking at the gluten content in the products.

22:25 And here we're going to compare how similar the words are,

22:28 and we're going to take out the sim the words that don't have um

22:33 A lot of information,

22:34 for instance,

22:34 drafts of these

22:38 um contains a lot of information per year,

22:39 and we're going to focus more on

22:41 um gluten-free,

22:42 very low gluten,

22:43 gluten content,

22:45 and we're going to compare

22:46 and we're going to take

22:47 an average of how dissimilar it is

22:51 between

22:52 the new regulations compared to what exists in the country as

22:55 a way of measuring how distant the two countries are.

23:00 Camil,

23:00 11 quick question here because this is relevant for,

23:03 for developing countries.

23:04 What do you do in terms of the distance where there is no regulation?

23:08 So,

23:09 um,

23:10 an African country is an exporter

23:13 and they don't have a gluten

23:15 regulation.

23:16 How,

23:16 how does this distance looks like?

23:19 Um,

23:22 So,

23:22 the,

23:23 we're gonna have an average per product per year,

23:26 I think,

23:27 per merger.

23:28 Um,

23:29 and I think if there's no regulations,

23:31 the distance is gonna be the maximum.

23:37 So 0

23:39 If they don't have any.

23:42 Uh,

23:42 it's gonna be one cause it's.

23:46 Yes,

23:46 1,

23:47 so 0 is gonna be 0 dissimilarity,

23:50 and then 1 is gonna be the maximum,

23:52 so maximum dissimilarity,

23:54 and that

23:55 implies then the adjustment costs are,

23:57 are larger.

23:59 Correct.

24:00 Right.

24:11 I just like to ask two quick questions also on this.

24:14 Um,

24:15 like,

24:16 um,

24:17 I guess like a lot of like TBTs also arise from

24:20 if the

24:21 kind of like technical thresholds will be different.

24:24 So I guess like if you compare the text,

24:26 it doesn't look like a big difference,

24:28 but in terms of production,

24:30 kind of like what it means for production can have like a huge difference.

24:33 Do you like somehow like is there like any technical like um

24:37 quantitative thresholds

24:39 mentioned

24:40 in the notifications,

24:41 did you try to use these differently or like make sense of it?

24:45 And the second one is,

24:47 I guess like the compliance

24:49 procedures matter a lot,

24:51 right,

24:51 in terms of conformity assessment,

24:53 um,

24:53 also like in terms of,

24:55 you know,

24:55 do you have to like use a notified body in the in the EU or like,

24:59 you know,

24:59 is it like recertification needed,

25:01 retesting,

25:02 is there like any way that you could

25:04 kind of like look at that in terms of the,

25:07 the impact in,

25:09 Um,

25:09 adaptation costs as well.

25:12 For compliance costs,

25:13 I would say.

25:16 Um,

25:16 so on the first question,

25:18 um,

25:20 We try to find um

25:23 Um,

25:23 regulation with the quantitative thresholds because at the beginning we

25:26 thought it was going to be easier to compare.

25:28 Um.

25:30 But the ones we found were more on pesticides,

25:32 which is more SBS and TBTs.

25:34 Um,

25:35 so if you have any idea of

25:38 One that could have a quantitivity threshold,

25:40 that would be great.

25:41 Um,

25:42 we tried to look for one we couldn't find.

25:44 Um,

25:45 we're still looking whether maybe in the textile industry,

25:48 there could be one if

25:49 maybe the content of textiles,

25:53 fibers could be specified.

25:54 Um,

25:55 so we're still looking at this.

25:57 Um,

25:58 and the second question,

25:59 um,

26:00 so in the beginning,

26:01 we were looking also at whether conformity assessment also have a different,

26:05 um,

26:06 effect

26:07 because the EU also signed a lot

26:08 of mutual recognition agreement with different countries.

26:11 Um,

26:12 So it was a bit two separate things,

26:15 um.

26:17 We're not sure whether that's actually part of um

26:21 Of adaptation costs or as you've mentioned,

26:23 is more compliance costs.

26:25 Um,

26:26 so maybe it would be an additional cost,

26:29 um,

26:29 to the TBTs,

26:31 um,

26:32 but maybe for a developing country which don't have testing facility,

26:35 facility in the countries,

26:36 that would be even harder.

26:39 To show,

26:40 not even to comply with the measures,

26:42 but to show that they are complying with the measures,

26:44 so that would be.

26:46 That,

26:46 that's a good point.

26:54 Thank you.

26:56 And sorry,

26:57 Camille,

26:57 sorry,

26:57 just,

26:57 just to,

26:58 to follow up on that,

26:59 so.

27:00 Um,

27:01 so,

27:01 one thing is what the regulation might be in the exporting country.

27:05 The other is what,

27:07 what exporters might be doing in that country,

27:09 right?

27:09 I mean,

27:09 there could be a,

27:10 you know,

27:11 there could be a long,

27:12 uh,

27:12 you know,

27:12 a,

27:13 a,

27:13 a wide difference between

27:16 In fact,

27:16 as Chavi was saying,

27:17 there could be no,

27:18 maybe there's no regulation,

27:20 uh,

27:21 say on coffee in Ethiopia.

27:22 I have no idea,

27:24 but,

27:24 but,

27:24 you know,

27:25 but there are,

27:26 you know,

27:26 Ethiopian exporters for,

27:28 for coffee that,

27:29 that comply with the DBTs.

27:32 So,

27:33 so,

27:33 so,

27:33 you know,

27:33 so this cost of adaptation is

27:36 I mean,

27:36 you know,

27:37 yeah,

27:38 why,

27:38 why is it on,

27:39 I mean,

27:39 it seems to me that this is like a,

27:41 an upper bound to,

27:42 uh,

27:42 to,

27:42 to the,

27:43 to the actual this because,

27:45 you know,

27:45 ultimately what you want is,

27:47 is for the exporters to comply,

27:48 right?

27:48 And so,

27:49 and so it's,

27:50 it's,

27:51 and again,

27:52 this could be,

27:52 this could be quite different

27:54 from,

27:55 from the actual regulation in the country,

27:57 right?

27:57 Uh,

27:58 cause,

27:59 yeah,

27:59 so,

28:00 yeah,

28:00 so,

28:00 so,

28:01 yeah,

28:01 I'm a bit,

28:02 yeah,

28:03 I'm,

28:03 I'm a bit confused.

28:04 Um.

28:07 So,

28:07 ideally,

28:08 we would have loved to have um firm to firm

28:11 data.

28:12 Um,

28:13 we couldn't find,

28:13 we only have firm to country.

28:15 Um

28:17 The only thing that was a bit reassuring is most of the time there seems to

28:23 To imports from

28:25 a few suppliers by country,

28:27 um.

28:29 Then the only thing is,

28:31 um,

28:31 if there is no regulation in the country.

28:35 Maybe

28:36 for Ethiopian

28:37 exporters,

28:37 the US is the big market,

28:39 so we're gonna decide to apply the US

28:41 um

28:43 regulations.

28:44 The only thing is,

28:45 I don't know how we would be able to know

28:48 whether that's the case or not.

28:50 Um,

28:51 maybe one thing we could look at is a few case studies,

28:54 whether that's the case in some countries.

28:57 And

28:58 So that's,

28:59 that's a good point,

29:00 um.

29:05 Yeah,

29:05 so here we're looking at whether

29:07 um

29:08 which regulations you have to follow to serve

29:10 the domestic market in your country,

29:13 uh,

29:13 but maybe you have firms that only exports

29:16 and

29:17 then have another

29:19 regulations.

29:20 Um,

29:22 But that's a good point.

29:23 I'm not sure how we could

29:24 check for this.

29:40 OK,

29:40 great.

29:41 Um,

29:43 So we're also gonna include um firm product origin

29:46 country fixed effects and sector origin country time fixed effects.

29:51 So here,

29:51 what we want to do is to exploit the time variation

29:55 uh in TBTs within firm product,

29:58 source countries,

29:59 um,

30:00 different import margin.

30:02 While controlling for sector country trends.

30:08 Um,

30:09 so,

30:09 if we look at the first,

30:10 um,

30:11 set of baseline,

30:12 um,

30:12 estimates,

30:13 so here we're gonna focus more on the

30:15 intensive margin.

30:16 So,

30:17 in the first,

30:18 um,

30:18 three columns,

30:19 you have the effects on the import value.

30:21 Um,

30:22 and here we can see,

30:23 um,

30:24 once the EU imposed a new TBT,

30:28 Um,

30:28 the import values of,

30:30 um,

30:30 towards EU country increased by around 1.8%.

30:35 Um,

30:36 if we distinguish a bit more between um how standards or how TBTs are,

30:41 um,

30:43 Deal with in specific countries or specific

30:46 um

30:47 products,

30:47 we can see.

30:49 But there is um quite a big increase

30:52 in the

30:54 um

30:56 In the import value of countries from which

30:59 either the EU has harmonized standards with or um

31:03 in the harmonized sectors within the EU.

31:06 So that increased by around 4.4%.

31:10 Um,

31:11 and here,

31:12 um,

31:12 which is quite surprising,

31:14 we see that if the EU only has mutual mutual recognition of standards

31:18 of the country,

31:19 the effect is negative.

31:21 Um,

31:22 and here it's mostly within the EU and here we found,

31:25 um,

31:25 there is a report by the EU Commission

31:28 that mentioned that uh mutual recognition that might not

31:31 work that well

31:33 between EU members and that sometimes they just decide.

31:37 They either got um denied at customs or the project got the product got rejected,

31:42 so they don't use it anymore.

31:45 And in the last columns,

31:47 um,

31:48 sorry,

31:48 in the third column,

31:49 we're using our,

31:50 um,

31:50 dissimilarity index.

31:52 And here we can see that if the standard deviation,

31:56 if the regulatory dissimilarity index increased by 1 standard deviation,

32:01 um,

32:02 the import value is gonna decrease by around 2.7%.

32:07 If we look at the 2nd part of the table here,

32:10 we're looking at the effect on import price.

32:14 And here we can see the import price increased,

32:17 um.

32:18 Um,

32:19 regardless of the

32:21 origins,

32:23 um.

32:25 Um,

32:25 due to the increasing costs or to the adaptation

32:28 costs of the production to the new standards.

32:34 If now we look more at um

32:37 A proxy of um supply chain relocation.

32:40 So here,

32:41 if we could focus on the 1st 3 columns,

32:44 that's the effect on quantity.

32:45 And here again,

32:46 we can see following the imposition of a TBT

32:49 we have an increase in quantity imported from origins,

32:52 um.

32:54 That have that have a low dissimilarity

32:57 or

32:57 that are harmonized with the EU.

33:00 Um

33:02 Um,

33:03 which is,

33:03 um,

33:03 then what is quite interesting is we can see the probability to exit,

33:08 um,

33:09 a sourcing links,

33:10 um,

33:11 Increase um from,

33:14 for the non.

33:16 Um,

33:16 so from the origin.

33:18 From the non-EU origin or for the origin that

33:21 are non-harmonized or doesn't have mutual recognition by around

33:25 um

33:27 And

33:29 1.3% points.

33:33 Um,

33:34 and then if we look at the last,

33:36 um,

33:36 part of the table,

33:38 here we're looking at the probability to enter a new

33:41 or to create a new sourcing link with a new,

33:44 um,

33:44 origin.

33:45 And here we can see the probability

33:48 to enter a new

33:50 Um,

33:50 or to start importing from a new origin

33:53 that is located outside of the EU decreases

33:56 by 2.6% points,

33:58 while it increased

34:00 by around 2 point,

34:02 a bit more than 2% points from harmonized origin.

34:08 And again,

34:08 if we use our um

34:11 Regulatory distance or dissimilarity index

34:15 between countries,

34:16 one standard deviation increase in this

34:18 uh index

34:19 decrease the probability to start

34:21 um importing from a new.

34:24 Um,

34:25 country by around 2.4% points.

34:33 OK.

34:33 Um,

34:34 so we've run several robustness checks,

34:37 um,

34:38 So,

34:39 as someone mentioned

34:40 earlier,

34:41 there might be some um endogeneity concerns about

34:44 uh which products are targeted by the EU.

34:47 Um,

34:48 so here,

34:48 since we're looking at France

34:50 and the

34:52 trade policies,

34:53 uh EU matter,

34:54 normally all the EU members um should agree on the product before setting the TPTs.

35:00 Um,

35:01 however,

35:02 France might still play a role in

35:05 um

35:06 selecting the products.

35:07 So we instrumented the TBTs

35:09 um using

35:10 the,

35:11 sorry,

35:11 we instrumented the EU TBTs using the TBTs um imposed by the US,

35:16 um.

35:18 To control for any

35:20 um

35:21 role played by the tariff,

35:22 we also control for this,

35:23 and um as an additional measure.

35:27 On top of just having a dummy equal to 1,

35:30 if there is a TPT or equal to 0,

35:31 otherwise,

35:32 we also control for uh or use the number of TPTs per product that is imposed.

35:38 Um

35:41 And then we wanted to go a bit deeper in trying to understand

35:45 um where this is coming from and what could explain

35:48 um the results we observed.

35:50 So,

35:50 the first one we think we had in mind is whether um

35:55 if

35:55 the search costs could also,

35:57 could have or could amplify the effect,

35:59 the tri diversion effect that we observed.

36:02 So we use the product stickiness measure that has been developed by Me and Caraus.

36:08 And here we find that links associated to products.

36:12 With higher relationship investment,

36:14 um,

36:17 Investments are less likely to be dropped.

36:20 Um,

36:20 secondly,

36:21 we also look at

36:22 whether the import reliance,

36:24 if you rely,

36:25 um,

36:25 a lot on a specific supplier,

36:27 whether that could also have an effect.

36:29 Um,

36:31 And

36:31 um here we use the France's initial reliance on a specific supplier

36:37 as a way of practicing for bargaining power,

36:40 and we found that

36:42 price pass through depends on these bargaining powers

36:44 and buyers heavily reliant on the source.

36:49 Face larger price rises and steeper quantity reductions.

36:54 And the last um

36:56 The angle we're looking at is whether

37:00 the scale of the finer good producers

37:02 also have an effect.

37:04 And here we found that larger firms,

37:07 um,

37:07 observe that larger firms are the ones that divert more.

37:10 And

37:12 we think it's because larger firms have a larger incentive to resume.

37:17 Um,

37:18 search to find a better match after the TBT is introduced.

37:24 Um,

37:25 so to sum up this first,

37:27 um,

37:27 paper,

37:27 what we found is,

37:28 uh,

37:29 GBTs do cause a trade diversion towards suppliers that face

37:33 lower adaptation costs.

37:36 Um,

37:36 we found also an impact.

37:37 They also have an impact on existing GBCs,

37:40 so the probability to enter a new market,

37:42 decreases by 2.6% points.

37:45 The exit increase from non-EU origin.

37:48 Um,

37:49 and also at the same time it strongly encourage

37:52 entry from harmonized or lower adaptation cost origins.

37:57 And we found that the drivers of um

38:01 The drivers of this diversions are explained by adaptation costs,

38:06 um,

38:06 switching switching costs,

38:08 bargaining power,

38:09 and the size of

38:10 the importer.

38:14 I don't know if

38:16 there is any question.

38:16 We have 5 minutes.

38:17 Uh,

38:18 we are in the half of the,

38:20 of the seminar,

38:23 um.

38:25 I wanted to understand better this mutual recognition effect,

38:28 right?

38:28 And how,

38:29 and how it overlaps as well with the distance,

38:32 right?

38:32 Because you would expect that.

38:34 Modal recognition and

38:37 And distancing regulator,

38:39 uh,

38:39 regulation are very correlated,

38:40 right?

38:41 You'll have mutual recognition when

38:43 you have similar standards,

38:45 no,

38:46 um,

38:47 so I wonder if,

38:48 if the results that you're finding were

38:50 almost,

38:51 I think I,

38:51 I read them,

38:53 uh,

38:53 as in even increasing farther trade diversion

38:57 may be related to the fact that it's,

38:59 it's correlated with your,

39:01 your distance measure.

39:05 Um,

39:09 So for

39:10 uh mutual recognition for the,

39:11 at least for the EU,

39:13 the we only found um

39:17 But the US mutual recognition of standards with

39:20 within the EU.

39:22 So we found that the EU has um harmonized standards with

39:26 I think 5 or 6 countries,

39:28 um,

39:29 but you only have mutual recognition

39:31 within the EU,

39:33 um.

39:35 So that's only within the EU.

39:37 Um,

39:37 and then we have,

39:39 um,

39:40 so in the non

39:41 In these sectors,

39:42 um,

39:43 some of the EU,

39:45 so you have TBT from the EU,

39:47 but then in these sectors,

39:48 you also have,

39:49 um,

39:50 TBTs from

39:51 the member countries itself.

39:53 So,

39:53 for instance,

39:53 in some products,

39:54 Germany is gonna have a TBT.

39:57 So that's where the distance is gonna be,

39:59 um,

40:00 in between.

40:01 Um,

40:02 but I could look a bit more,

40:03 um,

40:04 at the

40:07 If there's a link with the distance or

40:09 if there's a high correlation between mutual recognition and

40:13 Um,

40:13 and the distance measure.

40:18 Any,

40:19 any other questions,

40:20 Phillip?

40:20 You had your hand up.

40:22 Yeah,

40:22 I'm also like still on the regulatory distance point.

40:25 Um,

40:25 I was like wondering,

40:27 so like you base it on the

40:29 TBT notification document,

40:33 I assume,

40:33 but like not on the underlying

40:35 like legal text,

40:37 right,

40:37 which would provide more details and usually the TBT notification is just like a

40:42 summary which might

40:44 These are kind of a critical kind of like elements,

40:46 of course,

40:46 but I was wondering whether you you experimented with that.

40:49 And the second question is,

40:51 I mean,

40:51 you also have like the TB like the trade concerns that are that are raised

40:56 related to TBTs,

40:58 which might also give you like an indication of

41:00 if a lot of countries complain saying,

41:02 hey,

41:02 this is really bad,

41:03 this is really different from,

41:05 you know,

41:06 how it's like generally practiced in other parts of the world,

41:08 like,

41:10 You might have like an indication of the regulatory

41:13 distance as well and the seminarity,

41:16 so it's like anything that you like I tried out and kind of like

41:20 matched somehow the number of trade concerns raised by countries with

41:25 the regulatory disseminarity.

41:30 Um

41:32 So that's

41:33 really good points.

41:34 On the first one,

41:35 we haven't,

41:35 we use the TBT itself,

41:37 so we haven't checked within the documents.

41:41 Um,

41:44 I think we could do it for the countries,

41:46 but,

41:46 or,

41:47 or maybe not,

41:48 maybe we don't need um

41:51 Cause I guess we would need to um translate the document in one

41:55 language to be able to compare.

41:57 But that's something we could do,

41:58 but I guess maybe we're gonna lose some information into the translation,

42:02 but that's something we haven't looked at.

42:03 So maybe that's um

42:05 one option we could look at whether

42:07 to have a bit more

42:09 um

42:10 information.

42:11 And on the second point,

42:14 um,

42:16 We haven't checked whether we could match with the trade concerns,

42:19 um.

42:23 Well that's a good point because that would also be a bilateral measure,

42:26 so maybe some specific countries are going to complain about specific measures,

42:31 um.

42:32 So that could also be a way of saying this is really different from us,

42:36 um.

42:38 And we could know this specific group of countries

42:41 complain more about it rather than other countries that did not complain.

42:44 So we could look,

42:45 we haven't done it,

42:46 um,

42:46 but that's a good point.

42:47 Thank you.

42:54 If there are no more questions,

42:56 I suggest that we move to

42:58 To the second paper,

42:59 so we have time enough for discussion.

43:04 OK.

43:04 Thank you so much.

43:06 Um,

43:06 so now I'm gonna move to the second paper which looks more at,

43:10 um,

43:10 into what's the impact of voluntary standards.

43:14 Um,

43:15 so the idea of this paper and the motivation was,

43:17 um,

43:18 environmental awareness,

43:19 um,

43:19 has been increasing over,

43:21 over the past years.

43:23 Environmental impact of product is becoming more now

43:26 an important purchase criterion for many customers,

43:29 and as a results,

43:30 we observed that firms

43:32 are increasingly implementing voluntary environmental standards.

43:36 And,

43:37 these standards,

43:38 um,

43:39 help organizations identify,

43:41 manage,

43:41 monitor,

43:42 and control their,

43:43 um,

43:44 environmental issues.

43:45 And just to give you an example,

43:47 um,

43:47 and

43:48 Um,

43:49 how you can know,

43:50 uh,

43:51 where a firm is certified and how

43:53 you could be influenced,

43:54 um,

43:55 in your day to day life.

43:56 Um,

43:57 so that's an example of a truck that was,

43:59 uh,

44:00 parked outside of the university in Geneva

44:02 with,

44:03 um,

44:04 which was clearly marked,

44:05 but they are,

44:06 um,

44:06 certified to these three,

44:08 different ISO certification.

44:13 Um,

44:14 So to give you a bit of an idea of how um this stand um works.

44:19 So this stent is aimed

44:20 at helping the organization to reduce its environmental

44:23 impact and increase its um operating efficiency.

44:27 So we're gonna have um several steps.

44:29 The first one is we're gonna review the organization's environmental goals.

44:33 Um,

44:34 they're gonna analyze its environmental impacts and legal requirements,

44:38 and um then we're gonna be able to set environmental

44:41 objectives and targets um to reduce um its environmental impact.

44:46 They're gonna establish programs to meet uh the different objectives.

44:50 And finally,

44:51 we're going to ensure that the

44:52 employees' environmental awareness and competence is um

44:55 good enough.

44:58 So the idea of this paper is um

45:01 to understand what are the firm's motivations behind the certification.

45:04 So why would a firm decide to get certified even

45:07 if it comes at a cost and it's a voluntary

45:10 um certification.

45:13 And secondly,

45:14 what are the effects of voluntary certifications on firms' export performance?

45:19 So what's the effect um once the firms get certified?

45:23 So,

45:23 the main contributions are this um paper of freefold.

45:27 So the first one is um the first paper

45:29 to study the impact of of environmental cations on

45:34 both the extensive and intensive um firms.

45:38 And margins.

45:41 I'm gonna propose both an instrumental variable

45:44 strategy and a propensity score matching

45:47 um to account for the endogenous nature of the certification.

45:51 And I'm gonna investigate the differential effect across this nation,

45:56 countries,

45:57 products,

45:57 and firms

45:59 characteristics.

46:01 Um,

46:02 so for this,

46:03 um,

46:03 paper,

46:04 again,

46:04 I'm gonna use,

46:05 um,

46:06 French firm level data.

46:07 So I'm gonna have information on,

46:09 um,

46:10 exports at the firm year,

46:12 products,

46:12 and destination level.

46:14 Um,

46:15 I'm also gonna use administrative data.

46:17 So I'm gonna have,

46:18 um,

46:19 information

46:20 level based on tax reports.

46:23 And,

46:23 um,

46:24 I'm gonna have information that is coming from a more environmental data sets

46:29 which,

46:30 um,

46:31 collects information on the investment and studies made by the firms.

46:35 Towards protecting the environment

46:37 and it includes um the implementation of an environmental management system,

46:43 which is at the plant level.

46:45 So I'm gonna know whether the firm is certified or not.

46:50 Um,

46:51 in the data,

46:52 I'm gonna have information whether the firm is certified or not,

46:55 and it's going to be whether the firm is certified to

46:58 um one of its three standards.

47:00 So even the,

47:01 even the international one,

47:03 so the ISO 14001,

47:05 or some sort of the European equivalent

47:08 or

47:09 French equivalent.

47:11 Um

47:13 So nowadays there are more than 3000.

47:17 Certifications

47:20 in 171 countries.

47:23 Um,

47:25 As of 2020,

47:27 the more

47:29 European equivalent counted more than 3000 organizations and

47:36 12,

47:36 more than 12,000

47:38 plants.

47:40 And the French equivalent is um a model for the establishment and certifications

47:46 and to help firms get,

47:48 um,

47:48 reach one of these two other standards.

47:55 Um,

47:56 so just to give you an idea,

47:57 um,

47:58 so I,

47:59 I,

47:59 I'm gonna focus on the study period between 2002 and 2018.

48:03 Um,

48:04 and here,

48:04 as you can see,

48:05 the number of plants have been increasing

48:07 and over the period and,

48:09 um,

48:10 also the number of,

48:11 um,

48:12 certified

48:13 plants.

48:16 Um,

48:17 and here in my case to answer,

48:20 uh,

48:20 my research question,

48:21 I'm gonna,

48:22 um,

48:23 rely again on different,

48:25 I'm gonna look at the effect on different trade margins.

48:28 So I'm gonna look at um

48:30 the export value,

48:31 but also the quantity and the um

48:34 the effect on unit value.

48:36 And I'm also gonna look at the extensive margin and I'm gonna

48:39 look at the um effect on the probability to start exporting.

48:44 And here,

48:45 my um variable of interest

48:47 is going to be certification.

48:49 So,

48:50 the certification status of the firm is gonna be given at the plant level.

48:55 So,

48:55 I'm gonna have to aggregate this information at the firm level.

48:59 So,

48:59 I'm gonna use the share

49:01 of plants that are certified um

49:04 within the firm year combination.

49:07 Um,

49:08 I'm also gonna control whether the firm is certified to

49:12 any other certification of label in case there is any,

49:15 some sort of,

49:15 um,

49:16 learning effects

49:17 of if you're already certified to another label,

49:20 maybe it's gonna be easier to be certified to,

49:22 um,

49:23 to this environmental certification.

49:27 And I'm gonna control for,

49:29 uh,

49:29 firm uh fixed.

49:31 I'm gonna include firm fixed effects to control for

49:33 any factors that are specific to the firms,

49:36 and also gonna control for product destination year fixed effects.

49:39 So to control for any factors such as business cycle,

49:42 export demand shock,

49:44 um,

49:45 and the market

49:46 and,

49:46 uh,

49:47 competition conditions that the firm is facing.

49:50 So,

49:51 here again,

49:51 what I want to do with this um

49:53 identification strategy is to exploit the time variation

49:57 of certification in trade margins within firms that export the same product.

50:02 Um,

50:03 to the same destination,

50:04 um,

50:05 while controlling for any market conditions.

50:09 Um

50:14 Um,

50:14 so if we look at the first,

50:16 um,

50:17 um,

50:18 set of,

50:18 um,

50:19 baseline results,

50:20 um,

50:21 so here,

50:21 um,

50:22 what we found is following the certification,

50:25 um,

50:25 exports of the firms are going to increase by around 10%.

50:29 This increase,

50:30 uh,

50:31 in the exports.

50:33 Um,

50:34 it's gonna,

50:35 it's explained by both an increase in quantity but also an increase in

50:39 the unit value

50:41 of the firm.

50:43 And um

50:44 if we look more now into the extensive margin and

50:47 whether the firm is more likely to export afterwards,

50:50 um,

50:51 we also found a positive effect,

50:53 um.

50:55 So once the firms get certified,

50:56 it's more likely to start exporting to a new market.

51:02 And

51:04 So,

51:04 as I mentioned in the introduction,

51:06 um,

51:07 there's quite um big concern about the endogenous nature of the,

51:12 um,

51:12 firm.

51:14 Um,

51:14 so,

51:14 the first,

51:15 um,

51:16 endogene issue might be related to,

51:18 um,

51:18 omitted variable bias,

51:20 um,

51:21 which I think is quite,

51:22 um,

51:22 reduced by the use of,

51:24 um,

51:25 Of different um set of fixed effects.

51:29 Um,

51:30 And the results are also robust to the inclusion of

51:34 even more stringent

51:36 fixed effects to control for any composition effect within the film.

51:39 Um,

51:41 and.

51:43 So,

51:43 sorry,

51:43 but what if it,

51:45 this is simply,

51:46 you know,

51:46 certification is simply driven by the buyer

51:49 requiring the firm to be certified.

51:52 So,

51:53 you know,

51:53 I wanna buy from you,

51:55 but the only way to do that is for you to be certified.

51:58 And so that,

51:59 that's what explains.

52:01 So,

52:01 so,

52:01 it's,

52:02 it's basically,

52:02 you know,

52:02 reverse causality.

52:03 So,

52:04 you know,

52:05 I'm certified,

52:06 I,

52:06 I,

52:06 I am certified not because I want to export,

52:08 but,

52:09 sorry,

52:10 uh,

52:11 this is not about

52:12 certification,

52:13 improving the chances to,

52:15 to,

52:15 to export,

52:15 but actually,

52:17 you know,

52:17 uh,

52:18 you know,

52:18 uh yeah,

52:19 uh being required by the,

52:21 by the,

52:22 by the,

52:22 by the buyer,

52:23 by the importer.

52:24 The impor,

52:25 you know,

52:25 the only way I can export is if I,

52:28 if I'm,

52:28 if I'm certified.

52:32 Um

52:34 So,

52:34 I think that's a big point because um so last time I found a document from Tesco,

52:39 so a big retail store in the UK where they had

52:42 a big list of requirements that suppliers have to follow.

52:46 And at the end it was written,

52:47 if you are certified to this ISO standards,

52:51 you don't need to show that you are complying with all the other regulations.

52:55 So I think

52:56 Um,

52:57 that could be a way of,

52:59 um,

53:00 decrease some sort of

53:01 a way of decreasing,

53:03 um,

53:03 export costs,

53:05 um.

53:06 I don't know if that can be a requirement since it's a voluntary standards.

53:12 I guess officially cannot be said,

53:14 but maybe it's

53:15 implied,

53:17 and

53:19 So,

53:19 I guess one way of dealing with this that I'm doing is,

53:22 um,

53:23 and also to deal with selection

53:25 issue.

53:26 I'm comparing um

53:28 firms that are certified to firms that will

53:31 get certified

53:32 later during the study period.

53:34 Um,

53:35 so,

53:36 to control for the selection effective or I'm a big exporter,

53:40 I'm also gonna get certified,

53:41 um,

53:43 So I guess that's one way of dealing with it,

53:45 um.

53:47 The other way again,

53:48 that would be

53:49 if I had um

53:51 firm to firm data where I could control

53:54 for if a specific buyer has specific requirements.

53:58 Um,

53:59 I guess then in that case,

54:00 we would need all the suppliers from

54:03 a given,

54:04 all the buyers from a given country to also,

54:06 to want

54:07 them to be

54:09 um certified.

54:11 Um.

54:13 But I'm sure.

54:15 Do you have data on export destinations

54:17 for those firms because you could assume that

54:20 um

54:21 you could explore whether they are starting to export to other

54:25 destinations,

54:26 right?

54:27 Uh

54:28 yes,

54:28 still you could have the the issue of reverse causality because it may be

54:32 a multinational requiring the

54:34 the certification,

54:35 but

54:36 at least you could argue that

54:38 is is increasing the.

54:40 The,

54:40 the,

54:41 you know,

54:41 you're exporting to new destinations,

54:44 no.

54:45 Um,

54:45 so this I'm looking at whether,

54:47 um,

54:48 once you get certified,

54:49 you start exporting to new destination.

54:51 Um,

54:55 Yeah,

54:55 that's a good,

54:56 I could check.

55:07 Um,

55:08 yeah,

55:08 and the last point would be,

55:10 um,

55:10 so if the firm's export performance affects its decision to get certified.

55:16 Um,

55:17 so,

55:17 to deal with this,

55:18 um,

55:19 reverse causality,

55:20 I,

55:21 um,

55:22 I proposed two different methods.

55:24 So,

55:24 the first one is to rely on instrumental variable.

55:27 Um,

55:28 so,

55:28 in that case,

55:29 for each firm and year in the sample,

55:31 I'm going to compute the share of certified

55:33 plants that are located in the same

55:36 subnational region

55:38 in France producing the same products,

55:40 um,

55:41 as a way of controlling for,

55:44 um,

55:47 So,

55:47 any other um

55:50 variables that could uh influence your decisions to get certified

55:54 um other than the

55:59 The,

56:00 the

56:01 competition or the market conditions that you're facing.

56:05 And the second um solution I'm offering is to,

56:08 um,

56:09 to rely on or to employ 1 to 1

56:12 dynamic propensity score matching uh methods.

56:16 So in that case,

56:17 every,

56:17 for every firm

56:19 in the sample every year,

56:20 I'm gonna,

56:21 um,

56:24 I'm going to match this firm to a non-certified firm,

56:27 but which has the same probability to get certified.

56:32 Than the,

56:33 the treated firm.

56:37 Um,

56:37 and here,

56:37 if we look at the results using these two different methods,

56:41 um,

56:42 so the first four columns using is using the IV.

56:45 So here we can see the effects,

56:47 um,

56:47 are similar.

56:48 The design of the,

56:49 of the effects are similar to what we had in the baseline,

56:52 um,

56:52 but the magnitudes of the coefficients are,

56:55 um,

56:56 bigger,

56:57 which would suggest that um

57:00 Um,

57:01 The smaller firm would be the one getting certified.

57:05 Um,

57:05 and if we look at the,

57:07 um,

57:08 last three columns of the table here,

57:10 we can see the,

57:11 the results,

57:12 um,

57:12 using the propensity score matching.

57:14 And here we can see the results are,

57:16 um,

57:17 both the sign and the magnitude of the effect are really similar to what I had,

57:21 um,

57:21 in the baseline.

57:24 Um,

57:24 so I ran a lot,

57:25 quite a lot of,

57:26 um,

57:26 robustness checks.

57:28 Um,

57:30 Um,

57:31 so,

57:32 instead of,

57:32 um,

57:33 so I use,

57:33 uh,

57:34 instead of using the share of certified,

57:35 I use a dummy.

57:36 I also look at the effect on sales and export of sales,

57:40 and I found,

57:40 um,

57:41 also a positive effect on sales.

57:44 Um,

57:44 so I use the results also robust to

57:47 use another additional estimator to add different controls,

57:51 um.

57:53 Different fix effects,

57:54 etc.

57:56 And the last thing I wanted to look at and um is to look at whether

58:00 um what could be the potential drivers of this um certification.

58:04 So what would a firm,

58:06 what would be the firm's motivations behind the certification.

58:10 So the first thing I'm looking at is whether the effect

58:12 is different depending on the product the firm is exporting.

58:17 And here I use um the best classification

58:19 to distinguish between the different types of products.

58:22 And here I found um when we distinguish between intermediate and final goods,

58:27 um,

58:28 that the,

58:30 the quantity,

58:31 um,

58:31 exported increase,

58:33 um,

58:34 only for final goods,

58:35 uh,

58:36 which suggests that environmental supplication seems to

58:38 matter more for finer consumers than for,

58:40 um,

58:42 Um,

58:42 intermediate,

58:43 um,

58:44 buyers.

58:46 Um,

58:46 the second thing I was interested in is whether the effect is

58:49 different depending on where or to which market you're selling to.

58:53 So,

58:53 I look at the effect,

58:55 um,

58:55 first by the level of development.

58:57 Um,

58:58 and here,

58:58 the richer the country,

59:00 the larger the effect.

59:01 Um,

59:02 I also look at whether

59:04 if you have,

59:05 um,

59:05 as a way of measuring for environmental

59:07 awareness or environmental consciousness in the,

59:10 in the country.

59:11 I look at whether the EU has signed um

59:14 a trade agreement,

59:15 which includes environmental provisions

59:18 for those countries.

59:19 Um,

59:20 and here I observed that the increase in export volumes,

59:22 um,

59:24 Is quite similar,

59:25 but the price premium,

59:26 sorry,

59:26 um,

59:27 is observed.

59:29 And

59:30 with these countries.

59:32 And

59:33 And then as a last proxy for this uh environmental awareness,

59:37 I look at um

59:38 how big green imports represents in the total imports basket of the product.

59:44 And here I found that um certified firms are able

59:47 to charge a higher price to this um destination.

59:51 And the last thing I was,

59:53 I looked at and I'm interested in is whether

59:57 the effect is different depending on the firm.

59:59 So I use um

1:00:01 Value added per worker as a proxy for the firm productivity,

1:00:06 and here I found that larger firms are the ones

1:00:08 able to charge a higher price premium to consumers.

1:00:13 Camille,

1:00:13 can,

1:00:13 can we may perhaps before you go to the,

1:00:15 the conclusions,

1:00:16 explore the,

1:00:18 theterogeneity at the levels of development,

1:00:21 um,

1:00:22 which may be very relevant because

1:00:25 In a way

1:00:26 you would expect that if,

1:00:28 if

1:00:30 kind of the,

1:00:30 the,

1:00:31 the channel of the standard is,

1:00:32 is through signaling.

1:00:34 That some of the returns of that should be larger for developing economies,

1:00:38 right?

1:00:39 The cost may be larger,

1:00:40 but at least

1:00:41 the sort of the returns as well,

1:00:43 uh,

1:00:44 to export,

1:00:45 which should be larger,

1:00:47 is that not what you find?

1:00:49 Um,

1:00:50 so when I,

1:00:50 when I,

1:00:51 uh,

1:00:51 distinguish between developing and least developed economies,

1:00:55 um,

1:00:56 I found that the exports decreases.

1:00:59 Um,

1:00:59 the exports to least developed economies,

1:01:01 um,

1:01:02 decreases,

1:01:03 uh,

1:01:03 but increase for developing,

1:01:05 developing economies.

1:01:06 Um,

1:01:09 OK,

1:01:09 so,

1:01:09 so it's the mirror you're looking at the

1:01:10 exports to developing countries and to developed.

1:01:13 Oh,

1:01:13 OK.

1:01:14 All right,

1:01:14 right,

1:01:15 because I think,

1:01:15 um,

1:01:16 so I think if that would be the other run,

1:01:18 so if it would be exporters in developing countries,

1:01:21 I think that could be a really

1:01:23 helpful for them to get certified,

1:01:25 um.

1:01:27 And to export to more developed economies,

1:01:29 I think that could be.

1:01:31 Useful.

1:01:33 So the idea here of the mirror is that it

1:01:35 should be because the developing country markets are less demanding.

1:01:39 Uh,

1:01:40 you know,

1:01:41 the effects should be

1:01:42 lower

1:01:43 than for other

1:01:45 advanced countries.

1:01:46 OK,

1:01:46 so either there is less demand or there are

1:01:49 maybe a

1:01:51 Lower willingness to pay for a higher price for this

1:01:54 more environmentally friendly products in

1:01:57 And

1:01:58 more developing countries.

1:02:05 So maybe I can sum up this paper.

1:02:08 So what I found is environmental certification leads to an increase

1:02:13 in the export value by around 10%,

1:02:15 which is explained by a price premium of

1:02:19 a bit more than 2% and an increase in quantity exported by 9.6%.

1:02:24 So environmental certification seems to act as a quality signal for um

1:02:29 for these products,

1:02:31 for these firms.

1:02:32 Uh,

1:02:32 but it also creates new economic opportunities for certified firms.

1:02:37 So the probability to exports increased by 1.1% following the certification.

1:02:43 And the results are heterogeneous across destinations,

1:02:47 products,

1:02:47 and

1:02:48 firms.

1:02:51 Thank you very much.

1:02:54 Thank you,

1:02:54 Camille.

1:02:55 Uh,

1:02:55 let me open to

1:02:57 the questions.

1:03:00 Please feel free to ask.

1:03:04 Any questions?

1:03:08 Marilla.

1:03:13 Hi Camille,

1:03:13 thank you so much.

1:03:14 It's just,

1:03:15 I mean,

1:03:16 it's amazing work,

1:03:17 very,

1:03:17 very fascinating,

1:03:18 lots of interesting things,

1:03:19 uh,

1:03:20 that we've seen,

1:03:21 and

1:03:22 I just have a couple of uh questions,

1:03:24 like going back to your first presentation.

1:03:27 I was really interested in this regulatory dissimilarity index,

1:03:32 and I was,

1:03:33 it seemed to me if I paid enough attention that it stopped in 2019.

1:03:38 I was wondering if it could be expanded

1:03:41 and why did you stop in 2019.

1:03:43 Does,

1:03:44 does anything change in the.

1:03:46 Alright,

1:03:47 so just curiosity for that,

1:03:49 uh,

1:03:50 now second,

1:03:50 you said,

1:03:51 um,

1:03:51 that you showed us the differences in certification

1:03:56 impacting exports to developed and developing countries.

1:03:59 Uh,

1:03:59 yes,

1:03:59 the coefficient for developing countries was a little bit smaller,

1:04:04 but,

1:04:04 um,

1:04:05 they kind of look pretty much the same to me,

1:04:07 so I'm not sure if we can really speak about heterogeneity.

1:04:10 And another,

1:04:11 but again,

1:04:12 um,

1:04:12 you can clarify if I didn't pay attention to the,

1:04:15 to the,

1:04:15 the numbers,

1:04:16 uh,

1:04:17 um,

1:04:18 enough,

1:04:19 but,

1:04:19 um,

1:04:20 and then the final thing you just said

1:04:22 that the probability of exporting increases by 1%.

1:04:25 I was a bit

1:04:27 kind of disappointing.

1:04:28 I would disappointed.

1:04:29 I would have expected a bigger result assuming

1:04:32 that it's a costly process to certify and,

1:04:35 and so on.

1:04:36 And regardless of what's driven,

1:04:39 what's driving what,

1:04:40 you know,

1:04:40 the,

1:04:40 the

1:04:41 willingness to exports driving the need to certification or,

1:04:44 or the other way around,

1:04:45 it seemed like a very small effect,

1:04:47 uh,

1:04:48 for a supposedly significant effort.

1:04:51 I was wondering what's,

1:04:52 what's your take on this,

1:04:53 this probability of 1.4% increase in exports seemed,

1:04:56 seemed low to me.

1:04:58 Uh,

1:04:58 thanks so much.

1:05:02 Um,

1:05:02 so thank you very much.

1:05:03 On your first,

1:05:04 um,

1:05:05 question,

1:05:06 um,

1:05:07 we just stopped in 2019 cause that's when we started to

1:05:12 Um,

1:05:13 when we started collecting the data,

1:05:15 but we could easily update it and then add the new ones,

1:05:18 um.

1:05:20 It's just when that's when the all the data stops,

1:05:23 so I guess that's why,

1:05:24 um,

1:05:25 but that could easily be extended,

1:05:27 um.

1:05:28 On your second point,

1:05:29 um,

1:05:30 you're totally right.

1:05:31 The

1:05:32 It's more heterogeneous

1:05:34 with the least developed economies.

1:05:36 So the least developed

1:05:38 exports to the least developed economies,

1:05:40 the results were negative.

1:05:42 Um,

1:05:42 but indeed,

1:05:43 it was,

1:05:43 um,

1:05:44 positive for developing and developed.

1:05:46 Um,

1:05:47 so in between developing and developed,

1:05:49 I'm not,

1:05:49 um,

1:05:50 I could check whether the coefficients are different,

1:05:52 but I'm not sure.

1:05:53 So,

1:05:53 I think,

1:05:53 um,

1:05:54 you're right.

1:05:54 So,

1:05:54 it was more,

1:05:55 um,

1:05:57 With respect to least developed economies,

1:06:00 that there was an um heterogeneity.

1:06:04 And um on your last point,

1:06:05 um,

1:06:06 so I agree,

1:06:07 I think,

1:06:08 um,

1:06:10 So I was also expecting a bigger results,

1:06:12 um.

1:06:13 Um,

1:06:14 so one

1:06:15 explanation I might have,

1:06:16 um,

1:06:17 is

1:06:18 the

1:06:20 It might be because French firms are already exports to

1:06:24 Quite a lot of um destination already even before getting certified.

1:06:28 So that might be one explanation.

1:06:30 Um,

1:06:31 because I think the last time I found some paper

1:06:34 um that we're looking,

1:06:35 I think in Colombia

1:06:38 and Argentina and there or Ecuador and Argentina and then they

1:06:41 were mostly uh finding an effect on the extensive margin.

1:06:45 So then maybe going back to what Chavi was mentioning,

1:06:47 maybe for more developing countries,

1:06:50 um,

1:06:51 The effect could be even bigger for firms and um allow firms to export even more

1:06:56 to more destinations and to reach more markets.

1:06:59 So I think

1:07:00 maybe one reason why the magnitude is quite small is because France is already

1:07:05 quite developed economies and

1:07:07 um maybe firms are already exporting to quite a lot of markets.

1:07:14 Thank you.

1:07:15 That makes

1:07:16 a lot of sense.

1:07:16 Sorry,

1:07:17 just to follow up,

1:07:17 Chavi,

1:07:18 and I,

1:07:18 I promise,

1:07:19 um,

1:07:20 I was wondering if the reason,

1:07:21 uh,

1:07:22 you focused,

1:07:22 uh,

1:07:23 on France,

1:07:24 uh,

1:07:24 well,

1:07:25 it has,

1:07:26 uh,

1:07:26 something to do with the data quality as well.

1:07:28 I'm sure you have research interests in France,

1:07:30 but

1:07:31 do you know if similar data exists for developing countries that would allow

1:07:36 someone that would allow you to replicate this work for developing countries,

1:07:40 or?

1:07:41 Oh,

1:07:41 it's really like,

1:07:42 you know,

1:07:42 the,

1:07:43 because I know that French data is,

1:07:44 is super good,

1:07:45 but are,

1:07:46 I don't know about the quality of

1:07:48 this specific information for developing countries,

1:07:50 would that be replicable for in other

1:07:53 contexts?

1:07:54 Um,

1:07:55 so the main reason was data availability.

1:07:58 Um,

1:07:58 it's quite nice to have.

1:08:00 Um,

1:08:01 so I had a meeting with the International Standard Organization,

1:08:04 which are also really interesting in knowing

1:08:07 more what's the effect for more developing economies.

1:08:09 Um,

1:08:11 So far,

1:08:11 I think it's quite hard to find information at the firm level

1:08:15 on whether a firm is certified or not.

1:08:18 Um,

1:08:19 So I was looking,

1:08:21 um,

1:08:21 so I don't know if

1:08:23 the accreditation bodies um in the countries would be willing

1:08:27 to share the information on which firms are certified or not.

1:08:30 Um,

1:08:31 this,

1:08:31 I don't know.

1:08:33 But indeed,

1:08:33 I think that would be really interesting.

1:08:35 Um,

1:08:36 so I think it must exist for Argentina and Ecuador because I saw it

1:08:41 for

1:08:42 Um,

1:08:43 but I think Argentina was another standards,

1:08:45 but I think might,

1:08:46 um,

1:08:47 exist for other countries.

1:08:48 Um,

1:08:49 but I guess the

1:08:53 The aggregate,

1:08:53 um,

1:08:54 yeah,

1:08:54 maybe some firms are willing to share because otherwise it would,

1:08:57 uh,

1:08:58 be,

1:08:58 I think,

1:08:58 a long process to check online whether a firm is certified or not,

1:09:01 um,

1:09:02 if you have to do it one by one.

1:09:05 Thank you.

1:09:09 Thank you.

1:09:10 Any more questions?

1:09:14 No,

1:09:14 Perhaps I have a,

1:09:15 a final question,

1:09:16 at least on my end,

1:09:17 um,

1:09:18 on the terogeneity across sectors.

1:09:20 Uh,

1:09:20 you look a little bit

1:09:22 for the first paper on intermediates,

1:09:24 but,

1:09:24 uh,

1:09:25 I was wondering for the,

1:09:27 for the second paper,

1:09:28 probably environmental standards.

1:09:31 Matter differently for different sectors and exports.

1:09:34 I,

1:09:34 I was wondering if you,

1:09:35 if you explore the differences across sectors of,

1:09:38 of French exporters.

1:09:41 Um,

1:09:42 so I start looking at,

1:09:44 um,

1:09:44 I wanted to know,

1:09:45 for instance,

1:09:45 if the sector is more concentrated,

1:09:48 maybe if the effect is different,

1:09:50 um,

1:09:51 So I started looking into this,

1:09:53 um,

1:09:54 Maybe I could check a bit more,

1:09:56 look into um whether

1:10:00 Depending on sector characteristic if the effect is different,

1:10:03 that's a good point.

1:10:04 I haven't checked it yet.

1:10:11 OK,

1:10:11 colleagues,

1:10:12 uh,

1:10:13 if there are no more questions,

1:10:15 um,

1:10:16 I think we,

1:10:17 we can close here.

1:10:19 Uh,

1:10:19 thank you very much,

1:10:20 uh,

1:10:21 Camille,

1:10:22 for a very inspiring

1:10:24 discussion,

1:10:25 and,

1:10:25 uh,

1:10:25 the next seminar will be in May 15th.

1:10:29 The standard sizes the government role in diffusion

1:10:31 of mass production techniques in the US.

1:10:33 So

1:10:34 please stay tuned.

1:10:36 Thank you.

1:10:36 Thanks everybody.

1:10:37 Thank you very much.

1:10:38 Thank you for having me.

showAllTimestamps
no
transcript
So welcome everybody. This is, I think, um, The 3rd of our seminar series for the World Development Report on Standards for Development. And uh today we have uh Camille Reberti I, I pronounce it correctly uh Camille is a postdoc. She has a postdoc position at the at the graduate uh institute in Geneva. And she's been working on On, um, standard related issues to, to trade, um, and, um, we asked her to do a bit of an unconventional presentation where she will present, um, uh, two papers, two different papers, so we will have a bit of a break from one to the other, but very much related to, uh, the impact of standards around trade. The first one is on protection, our protectionism, the effect of technical regulations on input sourcing. And the second one is about adopting voluntary standards. Uh, so it's two different topics but very much related to, to the impacts of of standards on trade. We let her. To present for 45 minutes, uh, please, if you have any questions, uh, clarification questions first, we can, you can do it or send it on the, put it on the text box, uh, but otherwise we'll might make sure that we have a good 15 minutes or more for Q&A at the end, OK? So thanks Camille for doing this and, uh, over to you. Um, thank you very much. So, good morning, everyone. Um, I'm really happy to be here and present, um, Today, so as um Chavi mentioned, today I'm gonna present two different papers. So the presentation, um, it's called Navigating Standards, The Impact of Technical Regulations and Voluntary Certifications for firms trade um strategies. So as I'm sure you're all aware, global trade is nowadays increasingly shaped by standards such as non-tariff measures and voluntary standards. These different measures can enhance product quality and consumer trust, but they may also create some barriers to trade, despite their either voluntary or non-discriminatory nature. So as a result, understanding how such rules and regulations affect firms. Decisions, behavior, input sourcing decisions, or more trade performance in general is essential. And um in this presentation, I'm gonna focus on, I'm going to present two different papers. So the first one is gonna be on the effect of uh technical biers to trade, and the second one is going to focus more on um voluntary standards. Um. So we're a bit They are a bit different, um, as one is mandatory, whereas the other one is voluntary, but I think, um, they sort of complement each other in, in a way of, um, understanding more globally how um different types of regulation works and affects them. Um, so I'm gonna start with the first paper that looks at the effects of technical virus trade, and then I'll move on to the second paper. So, we, we started working on um the first chapter. So, um, bear in mind these figures might change a bit now that uh with the new Trump tariff, but, um, at least up to recent years, um, we observed a large decrease in tariff protection. Um, all over the world and it's even more true when we distinguish between type of products. So if you look here at the blue line, you can see that's even more stringent for um intermediate good where the um, The tariff rate is even lower than for for the average product. On the other hand, if we look at a similar picture before technical barriers to trade instead, we can see that they have been over the same period, they have been growing. And again, if we look at the blue line, um, it's even more striking for intermediate goods. So here, um, you have the three different types of products and you can see the largest increase are the, the products that have been targeted the most by technicalvis trade or intermediate goods. Um, before going a bit, um, deeper into the presentation, just a quick reminder of what, um, technical barriers to trade is. Um, so, TBTs are all the technical standards that differs from international one and might, might therefore impact trade. Um, to give you a more concrete example, part of the new Green Deal, the EU imposed a TPT on batteries and accumulators. And where In the regulations, if you want to use batteries, you have the EU impose requirements on the The, uh, level of hazardous materials that have been using. The level of um Of recycled materials that I have been using, and as you can see in the pictures, the batteries are um Widely used in different types of products. So it could be for electric vehicles, it could be for photovoltaic, um, products, etc. So, um, one specific activities might have a more global effect on across different, um, supply chains. So, what we are interesting in this paper and what we want to look at is, um, so do TBTs cause a diversion of supply chains and what could be the drivers um of these GVC disruptions? So, unlike tariffs, standards are non-discriminatory, uh, but they might have different effects on trade. So, first, um, the TBTs might reduce information of information asymmetries. Um, either about the quality of the product or, um, about whether the product is safe to use. And in that case, it might thus increase international sourcing. However, on the other hand, um, TBTs codify different technology that, uh, foreign firms might not have access to or might have access to other higher costs than dome domestic firms. And in that case, we would observed a decrease in international sourcing. Um, so what we wanna do and what I wanna show you is, um, what could be the effect of a TBT. Um, so, the main contribution of this paper are threefold. Um, so first, to answer this question, we're gonna create, we created a new database that covers all the different, um, TBTs that have been notified, uh, between 1995 and 2019. We're going to use product trade flows to estimate the effect of a TBT on the share of EU import sourcing. And But, um, we also include a more vertical mechanisms which I would not go, um, too deep into details, um, for this plantation, but the idea that we have is we're gonna incorporate adaptation costs, um, to the TBT into a model of sourcing. Um, we're also going to create a new measure, um, using text analysis techniques. Uh, we're going to offer a new measure of adaptation cost, and we're going to rely on French custom data, um. To show evidence of the microeconomic mechanisms of sourcing decisions. Um, So to give you an idea of how it works, um, and of the institutional framework of WTO, um, sorry, of TBT, um, so at the multilateral, multilateral levels, um, all the WTO members have signed the TBT agreements in 1995, and in these agreements, they agreed to notify to the WTO all the technical standards that differs from the international ones. At a more regional level, um, the EU has signed several PTAs with deep integration clauses. And in that case we're going to be interested in whether the EU has signed trade agreements which includes a clause on either harmonization of TBT or mutual recognition of TBTs. And even within the EU there are two different regimes, so around 70% of the production in the EU is within a harmonized sector, where in that case all the standards and all the rules and regulations are harmonized across all the EU members. But you also have the 30% remaining of the EU production that is in the non-harmonized or partly harmonized sector, which in that case, only mutual recognition applies across the EU members. Um, so, as I mentioned in the introduction, um, we're gonna use, um, several types of, um, data. The first one is, um, the new database that we created on all the different TBTs that have been notified to the WTO. So we collected information on around 30,000, um, TBTs around, uh, between 1995 and 2019. Um, however, out of this, um, 30,000 TBTs, around 40. 8% have no um product information. So, we implemented the multi-step procedures to recover the product information. So either we're gonna use um different sources, so we're going to look into the indemned, the notifications, different official documents. Um, we're going to match the ICS to the HS codes, um, and in the last step, we're gonna use a machine learning algorithm. Um, to have, uh, using the desk text descriptions, um, to match a certain notification to a certain, um, product information. And, then on the more regional level, we're going to use the World Bank database on provincial trade agreements and um. To have information on um DBT clauses that are included in the trade agreements. And for within the EU we're gonna use um data from the EU Commission that give us the list of non-harmonized sectors to distinguish between the products that are fully harmonized and the products that are not fully harmonized within the EU. Um, and finally, we're gonna use, um, French firm level data, um, which allow us to have, uh, to use a firm level definition of intermediate goods. So, we're gonna define an intermediate good as a good that is imported, um, And that is um in a different HS codes than the core business of the firm. Um, with the idea of, um, A different product may be an input for a firm, but might, might also be, um, a main export for the firm. So, we wanted to distinguish between a firm that is going to import, um, specific products for its, uh, for its production versus, um, a firm that is going to use this product as its main exports. Um, So, if we look at the aggregate level, um, what we can see is, um, so if we look at the effects of a TBT imposed by the EU on the share of EU that is sourced from within the EU. We can see activity increases the share of imports that is source within the EU. However, if we look at the effect for intermediate goods, uh, we observe that this is only 1/4 of the effects. Um, so it suggests that sourcing links are harder to substitute for intermediate goods. Um, so just one quick slide on the theoretical mechanisms and what we have in mind. Um, so we decided to add, um, the introduction of a TBT into a source model. And here, what we have in mind is a given TBT is going to require um suppliers to adapt their production. So if um A given country or a given supplier still want to export to the EU. They're going to have to adapt that production to make sure that they fulfill the new requirements imposed by the EU. And um we think this adaptation cost is going to be country specific and it's going to depend on how similar or dissimilar your production or your, or the regulation in your country or from the ones in the EU. And um this adaptation cost is going to affect the input prices directly through an increase in the unit cost, but also indirectly through an outside option. Um, so what we have in mind with this, um, vertical mechanisms is, um, So if we assume that um supply chains were previously located in Country A, um, now we're gonna be interested in whether um the country in you, in which you imports from. It's gonna change um following the introduction of a DBT. Um, so if the wage in country A, um, that is now taking into account, um, it uh, so our adaptation costs to the TBT is still lower than the wage, um, in Country B. We, um, we expect no change in the supplier origin, but we're still gonna, um, um. Still think they're gonna be an effect on the prices. Um, however, if now the wage, uh, that includes the adaptation costs in Country B becomes smaller, um, than the one in Country A, we now, um, accept, um, expect, expect, uh, ships from suppliers from Country A to Country B. Um, Um, so, to answer this question, Um, the research designs we're going to have is, so first we're going to look at different firms' import margin. So we're going to look at import expenditure and import prices. And on the more supply chains reallocation side effects, we're also going to look at the effects on quantity, entry and exit. And the way we're going to do it is, so our first variable of interest is going to be TBT, which is going to be equal to one, if the EU introduced a new TBT on a given product at a given time. And then we're gonna look at um Whether this effect is different depending on the adaptation cost. So it's a Um, TPT. We're gonna measure our EA TPT in two different ways. So the first one, the first way we're going to do it is to, yes. I mean, sorry, sorry, can, can we ask questions? Sorry, I, I'm, I'm not sure, yeah, yeah, OK, so, so sorry, just a quick point here. So, uh, so TDT. Um, you know, basically setting, uh, a TBT, the EU introducing a new TBT is, is I mean, there's a reason why they might do that, right? So, so, uh, so, you know, maybe they want to protect, maybe they want to, uh, I don't know, maybe there's a, uh, uh, uh, you know, safety scare and so they, they, they, so, So, so I guess, you know, are you dealing with, with this potential endogeneity or, or are you gonna tell us, uh, uh, you know, whether these, you know, DBTs tend to be introduced in cert at certain times in certain sectors? Uh, yeah, um, yeah, what, what's your, what's your take on that? I mean, so, this is like, you know, they're exhaustionously uh given, but, but clearly they're not, right? So, yeah, so. Um, so one way we have of dealing with this is, um, so sometimes when they notify the TPT to the WTO, they also give an objective of the measure, and we also look at whether the effect is still there if we control for either if the TPT as an objective of harmonization. Of, of trade enhancing, um, Um, and there we still have our negative effect of the other TBTs. Um, we try to look at, um, The different effects by different um objectives of the TBT, um. We also found, um, so I'm not sure I put it on the slides, but even if the objective is protection of the environment or, um, organic certification, something like this, um, the effect is also sometimes negative. Um, So, the one we looked at were more whether the objective was trade enhancing, harmonization, um, But we also found, I think when we looked at either uh either protection of the environment or consumer protection, I think it was also one objective, and there the effect was still negative. And, and sorry, just a, a related point here. So, so TPTs now are, are, are sort of dummy, dummy measures, but, but there could be You know, there could be, uh, you know, these changes could be quite different, right? Relative to what, uh, what the, what the standard was, uh, before the, this introduction, right? I mean, so, so it could be a big jump, jump or it could be a smaller, a smaller jump so that, so, so that, you know, the, the, the costs for compliance, the compliance caused by the, by the exporter, uh, could be, could be quite different. So, Yeah, so, uh, uh, yeah, are you, are you looking at that or, or that's maybe not a relevant margin? Yeah, what, what, what, what, what do you, what do you think? Um, so, in the baseline, we use a dummy. One robustness check that we do is to use account numbers. So if a product is targeted by 10 TBTs that might be different than if it's only 2. So that's one thing we do. Um, and then on the slide afterwards, I'll show you how we measure this adaptation cost. And the way we do it is to compare, um, the regulations, the newly regulation that is introduced versus, um, the regulations that is in place in the country, um. Maybe one additional thing we could do is to compare. In between the EU regulations, cause I think that's what you meant, so, um, Yeah, that's something we haven't looked at. So, we checked every new regulations, how different it is from what is applied, but not how different the new one is compared to the other one. Um, I think we tried to do it. So instead of, uh, comparing the new ones to the one that is in place, we also compare the, all the EU measures, including the new ones to everything. Um, and I think the results were quite similar. Or maybe we could look a bit more specific into changes in between the regulations, in between duties. OK. Um, So the first way we measure the adaptation cost is to compare um the effect the effect that TBT is gonna have um within the EU compared to outside of the EU. We have the idea of um The EU suppliers might have more um incentive to Uh, to comply with the new regulations because even if they only want to serve the domestic market, they have to, um, comply with the new regulations versus an extra EU suppliers, um, depending on how big the EU market represents, might not be willing to, um, implement the new standards. And the second way we're going to measure our adaptation cost is to look at how dissimilar the The regulation in the EU is compared to the regulation in the country you want to import from. So here to give you an example, um, we're gonna compare, so, let's say TBT3 is the new EU TBT that is imposed. And uh we want to compare it to um to how distance it is from the irrigation that is in place in Australia. So in that case, we're gonna compare the contents of the TBT of the EU TBT to the content of um the TBTs imposed in Australia. So, here we put below um an example. Um, so TB23 is now um is looking at the gluten content in the products. And here we're going to compare how similar the words are, and we're going to take out the sim the words that don't have um A lot of information, for instance, drafts of these um contains a lot of information per year, and we're going to focus more on um gluten-free, very low gluten, gluten content, and we're going to compare and we're going to take an average of how dissimilar it is between the new regulations compared to what exists in the country as a way of measuring how distant the two countries are. Camil, 11 quick question here because this is relevant for, for developing countries. What do you do in terms of the distance where there is no regulation? So, um, an African country is an exporter and they don't have a gluten regulation. How, how does this distance looks like? Um, So, the, we're gonna have an average per product per year, I think, per merger. Um, and I think if there's no regulations, the distance is gonna be the maximum. So 0 If they don't have any. Uh, it's gonna be one cause it's. Yes, 1, so 0 is gonna be 0 dissimilarity, and then 1 is gonna be the maximum, so maximum dissimilarity, and that implies then the adjustment costs are, are larger. Correct. Right. I just like to ask two quick questions also on this. Um, like, um, I guess like a lot of like TBTs also arise from if the kind of like technical thresholds will be different. So I guess like if you compare the text, it doesn't look like a big difference, but in terms of production, kind of like what it means for production can have like a huge difference. Do you like somehow like is there like any technical like um quantitative thresholds mentioned in the notifications, did you try to use these differently or like make sense of it? And the second one is, I guess like the compliance procedures matter a lot, right, in terms of conformity assessment, um, also like in terms of, you know, do you have to like use a notified body in the in the EU or like, you know, is it like recertification needed, retesting, is there like any way that you could kind of like look at that in terms of the, the impact in, Um, adaptation costs as well. For compliance costs, I would say. Um, so on the first question, um, We try to find um Um, regulation with the quantitative thresholds because at the beginning we thought it was going to be easier to compare. Um. But the ones we found were more on pesticides, which is more SBS and TBTs. Um, so if you have any idea of One that could have a quantitivity threshold, that would be great. Um, we tried to look for one we couldn't find. Um, we're still looking whether maybe in the textile industry, there could be one if maybe the content of textiles, fibers could be specified. Um, so we're still looking at this. Um, and the second question, um, so in the beginning, we were looking also at whether conformity assessment also have a different, um, effect because the EU also signed a lot of mutual recognition agreement with different countries. Um, So it was a bit two separate things, um. We're not sure whether that's actually part of um Of adaptation costs or as you've mentioned, is more compliance costs. Um, so maybe it would be an additional cost, um, to the TBTs, um, but maybe for a developing country which don't have testing facility, facility in the countries, that would be even harder. To show, not even to comply with the measures, but to show that they are complying with the measures, so that would be. That, that's a good point. Thank you. And sorry, Camille, sorry, just, just to, to follow up on that, so. Um, so, one thing is what the regulation might be in the exporting country. The other is what, what exporters might be doing in that country, right? I mean, there could be a, you know, there could be a long, uh, you know, a, a, a wide difference between In fact, as Chavi was saying, there could be no, maybe there's no regulation, uh, say on coffee in Ethiopia. I have no idea, but, but, you know, but there are, you know, Ethiopian exporters for, for coffee that, that comply with the DBTs. So, so, so, you know, so this cost of adaptation is I mean, you know, yeah, why, why is it on, I mean, it seems to me that this is like a, an upper bound to, uh, to, to the, to the actual this because, you know, ultimately what you want is, is for the exporters to comply, right? And so, and so it's, it's, and again, this could be, this could be quite different from, from the actual regulation in the country, right? Uh, cause, yeah, so, yeah, so, so, yeah, I'm a bit, yeah, I'm, I'm a bit confused. Um. So, ideally, we would have loved to have um firm to firm data. Um, we couldn't find, we only have firm to country. Um The only thing that was a bit reassuring is most of the time there seems to To imports from a few suppliers by country, um. Then the only thing is, um, if there is no regulation in the country. Maybe for Ethiopian exporters, the US is the big market, so we're gonna decide to apply the US um regulations. The only thing is, I don't know how we would be able to know whether that's the case or not. Um, maybe one thing we could look at is a few case studies, whether that's the case in some countries. And So that's, that's a good point, um. Yeah, so here we're looking at whether um which regulations you have to follow to serve the domestic market in your country, uh, but maybe you have firms that only exports and then have another regulations. Um, But that's a good point. I'm not sure how we could check for this. OK, great. Um, So we're also gonna include um firm product origin country fixed effects and sector origin country time fixed effects. So here, what we want to do is to exploit the time variation uh in TBTs within firm product, source countries, um, different import margin. While controlling for sector country trends. Um, so, if we look at the first, um, set of baseline, um, estimates, so here we're gonna focus more on the intensive margin. So, in the first, um, three columns, you have the effects on the import value. Um, and here we can see, um, once the EU imposed a new TBT, Um, the import values of, um, towards EU country increased by around 1.8%. Um, if we distinguish a bit more between um how standards or how TBTs are, um, Deal with in specific countries or specific um products, we can see. But there is um quite a big increase in the um In the import value of countries from which either the EU has harmonized standards with or um in the harmonized sectors within the EU. So that increased by around 4.4%. Um, and here, um, which is quite surprising, we see that if the EU only has mutual mutual recognition of standards of the country, the effect is negative. Um, and here it's mostly within the EU and here we found, um, there is a report by the EU Commission that mentioned that uh mutual recognition that might not work that well between EU members and that sometimes they just decide. They either got um denied at customs or the project got the product got rejected, so they don't use it anymore. And in the last columns, um, sorry, in the third column, we're using our, um, dissimilarity index. And here we can see that if the standard deviation, if the regulatory dissimilarity index increased by 1 standard deviation, um, the import value is gonna decrease by around 2.7%. If we look at the 2nd part of the table here, we're looking at the effect on import price. And here we can see the import price increased, um. Um, regardless of the origins, um. Um, due to the increasing costs or to the adaptation costs of the production to the new standards. If now we look more at um A proxy of um supply chain relocation. So here, if we could focus on the 1st 3 columns, that's the effect on quantity. And here again, we can see following the imposition of a TBT we have an increase in quantity imported from origins, um. That have that have a low dissimilarity or that are harmonized with the EU. Um Um, which is, um, then what is quite interesting is we can see the probability to exit, um, a sourcing links, um, Increase um from, for the non. Um, so from the origin. From the non-EU origin or for the origin that are non-harmonized or doesn't have mutual recognition by around um And 1.3% points. Um, and then if we look at the last, um, part of the table, here we're looking at the probability to enter a new or to create a new sourcing link with a new, um, origin. And here we can see the probability to enter a new Um, or to start importing from a new origin that is located outside of the EU decreases by 2.6% points, while it increased by around 2 point, a bit more than 2% points from harmonized origin. And again, if we use our um Regulatory distance or dissimilarity index between countries, one standard deviation increase in this uh index decrease the probability to start um importing from a new. Um, country by around 2.4% points. OK. Um, so we've run several robustness checks, um, So, as someone mentioned earlier, there might be some um endogeneity concerns about uh which products are targeted by the EU. Um, so here, since we're looking at France and the trade policies, uh EU matter, normally all the EU members um should agree on the product before setting the TPTs. Um, however, France might still play a role in um selecting the products. So we instrumented the TBTs um using the, sorry, we instrumented the EU TBTs using the TBTs um imposed by the US, um. To control for any um role played by the tariff, we also control for this, and um as an additional measure. On top of just having a dummy equal to 1, if there is a TPT or equal to 0, otherwise, we also control for uh or use the number of TPTs per product that is imposed. Um And then we wanted to go a bit deeper in trying to understand um where this is coming from and what could explain um the results we observed. So, the first one we think we had in mind is whether um if the search costs could also, could have or could amplify the effect, the tri diversion effect that we observed. So we use the product stickiness measure that has been developed by Me and Caraus. And here we find that links associated to products. With higher relationship investment, um, Investments are less likely to be dropped. Um, secondly, we also look at whether the import reliance, if you rely, um, a lot on a specific supplier, whether that could also have an effect. Um, And um here we use the France's initial reliance on a specific supplier as a way of practicing for bargaining power, and we found that price pass through depends on these bargaining powers and buyers heavily reliant on the source. Face larger price rises and steeper quantity reductions. And the last um The angle we're looking at is whether the scale of the finer good producers also have an effect. And here we found that larger firms, um, observe that larger firms are the ones that divert more. And we think it's because larger firms have a larger incentive to resume. Um, search to find a better match after the TBT is introduced. Um, so to sum up this first, um, paper, what we found is, uh, GBTs do cause a trade diversion towards suppliers that face lower adaptation costs. Um, we found also an impact. They also have an impact on existing GBCs, so the probability to enter a new market, decreases by 2.6% points. The exit increase from non-EU origin. Um, and also at the same time it strongly encourage entry from harmonized or lower adaptation cost origins. And we found that the drivers of um The drivers of this diversions are explained by adaptation costs, um, switching switching costs, bargaining power, and the size of the importer. I don't know if there is any question. We have 5 minutes. Uh, we are in the half of the, of the seminar, um. I wanted to understand better this mutual recognition effect, right? And how, and how it overlaps as well with the distance, right? Because you would expect that. Modal recognition and And distancing regulator, uh, regulation are very correlated, right? You'll have mutual recognition when you have similar standards, no, um, so I wonder if, if the results that you're finding were almost, I think I, I read them, uh, as in even increasing farther trade diversion may be related to the fact that it's, it's correlated with your, your distance measure. Um, So for uh mutual recognition for the, at least for the EU, the we only found um But the US mutual recognition of standards with within the EU. So we found that the EU has um harmonized standards with I think 5 or 6 countries, um, but you only have mutual recognition within the EU, um. So that's only within the EU. Um, and then we have, um, so in the non In these sectors, um, some of the EU, so you have TBT from the EU, but then in these sectors, you also have, um, TBTs from the member countries itself. So, for instance, in some products, Germany is gonna have a TBT. So that's where the distance is gonna be, um, in between. Um, but I could look a bit more, um, at the If there's a link with the distance or if there's a high correlation between mutual recognition and Um, and the distance measure. Any, any other questions, Phillip? You had your hand up. Yeah, I'm also like still on the regulatory distance point. Um, I was like wondering, so like you base it on the TBT notification document, I assume, but like not on the underlying like legal text, right, which would provide more details and usually the TBT notification is just like a summary which might These are kind of a critical kind of like elements, of course, but I was wondering whether you you experimented with that. And the second question is, I mean, you also have like the TB like the trade concerns that are that are raised related to TBTs, which might also give you like an indication of if a lot of countries complain saying, hey, this is really bad, this is really different from, you know, how it's like generally practiced in other parts of the world, like, You might have like an indication of the regulatory distance as well and the seminarity, so it's like anything that you like I tried out and kind of like matched somehow the number of trade concerns raised by countries with the regulatory disseminarity. Um So that's really good points. On the first one, we haven't, we use the TBT itself, so we haven't checked within the documents. Um, I think we could do it for the countries, but, or, or maybe not, maybe we don't need um Cause I guess we would need to um translate the document in one language to be able to compare. But that's something we could do, but I guess maybe we're gonna lose some information into the translation, but that's something we haven't looked at. So maybe that's um one option we could look at whether to have a bit more um information. And on the second point, um, We haven't checked whether we could match with the trade concerns, um. Well that's a good point because that would also be a bilateral measure, so maybe some specific countries are going to complain about specific measures, um. So that could also be a way of saying this is really different from us, um. And we could know this specific group of countries complain more about it rather than other countries that did not complain. So we could look, we haven't done it, um, but that's a good point. Thank you. If there are no more questions, I suggest that we move to To the second paper, so we have time enough for discussion. OK. Thank you so much. Um, so now I'm gonna move to the second paper which looks more at, um, into what's the impact of voluntary standards. Um, so the idea of this paper and the motivation was, um, environmental awareness, um, has been increasing over, over the past years. Environmental impact of product is becoming more now an important purchase criterion for many customers, and as a results, we observed that firms are increasingly implementing voluntary environmental standards. And, these standards, um, help organizations identify, manage, monitor, and control their, um, environmental issues. And just to give you an example, um, and Um, how you can know, uh, where a firm is certified and how you could be influenced, um, in your day to day life. Um, so that's an example of a truck that was, uh, parked outside of the university in Geneva with, um, which was clearly marked, but they are, um, certified to these three, different ISO certification. Um, So to give you a bit of an idea of how um this stand um works. So this stent is aimed at helping the organization to reduce its environmental impact and increase its um operating efficiency. So we're gonna have um several steps. The first one is we're gonna review the organization's environmental goals. Um, they're gonna analyze its environmental impacts and legal requirements, and um then we're gonna be able to set environmental objectives and targets um to reduce um its environmental impact. They're gonna establish programs to meet uh the different objectives. And finally, we're going to ensure that the employees' environmental awareness and competence is um good enough. So the idea of this paper is um to understand what are the firm's motivations behind the certification. So why would a firm decide to get certified even if it comes at a cost and it's a voluntary um certification. And secondly, what are the effects of voluntary certifications on firms' export performance? So what's the effect um once the firms get certified? So, the main contributions are this um paper of freefold. So the first one is um the first paper to study the impact of of environmental cations on both the extensive and intensive um firms. And margins. I'm gonna propose both an instrumental variable strategy and a propensity score matching um to account for the endogenous nature of the certification. And I'm gonna investigate the differential effect across this nation, countries, products, and firms characteristics. Um, so for this, um, paper, again, I'm gonna use, um, French firm level data. So I'm gonna have information on, um, exports at the firm year, products, and destination level. Um, I'm also gonna use administrative data. So I'm gonna have, um, information level based on tax reports. And, um, I'm gonna have information that is coming from a more environmental data sets which, um, collects information on the investment and studies made by the firms. Towards protecting the environment and it includes um the implementation of an environmental management system, which is at the plant level. So I'm gonna know whether the firm is certified or not. Um, in the data, I'm gonna have information whether the firm is certified or not, and it's going to be whether the firm is certified to um one of its three standards. So even the, even the international one, so the ISO 14001, or some sort of the European equivalent or French equivalent. Um So nowadays there are more than 3000. Certifications in 171 countries. Um, As of 2020, the more European equivalent counted more than 3000 organizations and 12, more than 12,000 plants. And the French equivalent is um a model for the establishment and certifications and to help firms get, um, reach one of these two other standards. Um, so just to give you an idea, um, so I, I, I'm gonna focus on the study period between 2002 and 2018. Um, and here, as you can see, the number of plants have been increasing and over the period and, um, also the number of, um, certified plants. Um, and here in my case to answer, uh, my research question, I'm gonna, um, rely again on different, I'm gonna look at the effect on different trade margins. So I'm gonna look at um the export value, but also the quantity and the um the effect on unit value. And I'm also gonna look at the extensive margin and I'm gonna look at the um effect on the probability to start exporting. And here, my um variable of interest is going to be certification. So, the certification status of the firm is gonna be given at the plant level. So, I'm gonna have to aggregate this information at the firm level. So, I'm gonna use the share of plants that are certified um within the firm year combination. Um, I'm also gonna control whether the firm is certified to any other certification of label in case there is any, some sort of, um, learning effects of if you're already certified to another label, maybe it's gonna be easier to be certified to, um, to this environmental certification. And I'm gonna control for, uh, firm uh fixed. I'm gonna include firm fixed effects to control for any factors that are specific to the firms, and also gonna control for product destination year fixed effects. So to control for any factors such as business cycle, export demand shock, um, and the market and, uh, competition conditions that the firm is facing. So, here again, what I want to do with this um identification strategy is to exploit the time variation of certification in trade margins within firms that export the same product. Um, to the same destination, um, while controlling for any market conditions. Um Um, so if we look at the first, um, um, set of, um, baseline results, um, so here, um, what we found is following the certification, um, exports of the firms are going to increase by around 10%. This increase, uh, in the exports. Um, it's gonna, it's explained by both an increase in quantity but also an increase in the unit value of the firm. And um if we look more now into the extensive margin and whether the firm is more likely to export afterwards, um, we also found a positive effect, um. So once the firms get certified, it's more likely to start exporting to a new market. And So, as I mentioned in the introduction, um, there's quite um big concern about the endogenous nature of the, um, firm. Um, so, the first, um, endogene issue might be related to, um, omitted variable bias, um, which I think is quite, um, reduced by the use of, um, Of different um set of fixed effects. Um, And the results are also robust to the inclusion of even more stringent fixed effects to control for any composition effect within the film. Um, and. So, sorry, but what if it, this is simply, you know, certification is simply driven by the buyer requiring the firm to be certified. So, you know, I wanna buy from you, but the only way to do that is for you to be certified. And so that, that's what explains. So, so, it's, it's basically, you know, reverse causality. So, you know, I'm certified, I, I, I am certified not because I want to export, but, sorry, uh, this is not about certification, improving the chances to, to, to export, but actually, you know, uh, you know, uh yeah, uh being required by the, by the, by the, by the buyer, by the importer. The impor, you know, the only way I can export is if I, if I'm, if I'm certified. Um So, I think that's a big point because um so last time I found a document from Tesco, so a big retail store in the UK where they had a big list of requirements that suppliers have to follow. And at the end it was written, if you are certified to this ISO standards, you don't need to show that you are complying with all the other regulations. So I think Um, that could be a way of, um, decrease some sort of a way of decreasing, um, export costs, um. I don't know if that can be a requirement since it's a voluntary standards. I guess officially cannot be said, but maybe it's implied, and So, I guess one way of dealing with this that I'm doing is, um, and also to deal with selection issue. I'm comparing um firms that are certified to firms that will get certified later during the study period. Um, so, to control for the selection effective or I'm a big exporter, I'm also gonna get certified, um, So I guess that's one way of dealing with it, um. The other way again, that would be if I had um firm to firm data where I could control for if a specific buyer has specific requirements. Um, I guess then in that case, we would need all the suppliers from a given, all the buyers from a given country to also, to want them to be um certified. Um. But I'm sure. Do you have data on export destinations for those firms because you could assume that um you could explore whether they are starting to export to other destinations, right? Uh yes, still you could have the the issue of reverse causality because it may be a multinational requiring the the certification, but at least you could argue that is is increasing the. The, the, you know, you're exporting to new destinations, no. Um, so this I'm looking at whether, um, once you get certified, you start exporting to new destination. Um, Yeah, that's a good, I could check. Um, yeah, and the last point would be, um, so if the firm's export performance affects its decision to get certified. Um, so, to deal with this, um, reverse causality, I, um, I proposed two different methods. So, the first one is to rely on instrumental variable. Um, so, in that case, for each firm and year in the sample, I'm going to compute the share of certified plants that are located in the same subnational region in France producing the same products, um, as a way of controlling for, um, So, any other um variables that could uh influence your decisions to get certified um other than the The, the competition or the market conditions that you're facing. And the second um solution I'm offering is to, um, to rely on or to employ 1 to 1 dynamic propensity score matching uh methods. So in that case, every, for every firm in the sample every year, I'm gonna, um, I'm going to match this firm to a non-certified firm, but which has the same probability to get certified. Than the, the treated firm. Um, and here, if we look at the results using these two different methods, um, so the first four columns using is using the IV. So here we can see the effects, um, are similar. The design of the, of the effects are similar to what we had in the baseline, um, but the magnitudes of the coefficients are, um, bigger, which would suggest that um Um, The smaller firm would be the one getting certified. Um, and if we look at the, um, last three columns of the table here, we can see the, the results, um, using the propensity score matching. And here we can see the results are, um, both the sign and the magnitude of the effect are really similar to what I had, um, in the baseline. Um, so I ran a lot, quite a lot of, um, robustness checks. Um, Um, so, instead of, um, so I use, uh, instead of using the share of certified, I use a dummy. I also look at the effect on sales and export of sales, and I found, um, also a positive effect on sales. Um, so I use the results also robust to use another additional estimator to add different controls, um. Different fix effects, etc. And the last thing I wanted to look at and um is to look at whether um what could be the potential drivers of this um certification. So what would a firm, what would be the firm's motivations behind the certification. So the first thing I'm looking at is whether the effect is different depending on the product the firm is exporting. And here I use um the best classification to distinguish between the different types of products. And here I found um when we distinguish between intermediate and final goods, um, that the, the quantity, um, exported increase, um, only for final goods, uh, which suggests that environmental supplication seems to matter more for finer consumers than for, um, Um, intermediate, um, buyers. Um, the second thing I was interested in is whether the effect is different depending on where or to which market you're selling to. So, I look at the effect, um, first by the level of development. Um, and here, the richer the country, the larger the effect. Um, I also look at whether if you have, um, as a way of measuring for environmental awareness or environmental consciousness in the, in the country. I look at whether the EU has signed um a trade agreement, which includes environmental provisions for those countries. Um, and here I observed that the increase in export volumes, um, Is quite similar, but the price premium, sorry, um, is observed. And with these countries. And And then as a last proxy for this uh environmental awareness, I look at um how big green imports represents in the total imports basket of the product. And here I found that um certified firms are able to charge a higher price to this um destination. And the last thing I was, I looked at and I'm interested in is whether the effect is different depending on the firm. So I use um Value added per worker as a proxy for the firm productivity, and here I found that larger firms are the ones able to charge a higher price premium to consumers. Camille, can, can we may perhaps before you go to the, the conclusions, explore the, theterogeneity at the levels of development, um, which may be very relevant because In a way you would expect that if, if kind of the, the, the channel of the standard is, is through signaling. That some of the returns of that should be larger for developing economies, right? The cost may be larger, but at least the sort of the returns as well, uh, to export, which should be larger, is that not what you find? Um, so when I, when I, uh, distinguish between developing and least developed economies, um, I found that the exports decreases. Um, the exports to least developed economies, um, decreases, uh, but increase for developing, developing economies. Um, OK, so, so it's the mirror you're looking at the exports to developing countries and to developed. Oh, OK. All right, right, because I think, um, so I think if that would be the other run, so if it would be exporters in developing countries, I think that could be a really helpful for them to get certified, um. And to export to more developed economies, I think that could be. Useful. So the idea here of the mirror is that it should be because the developing country markets are less demanding. Uh, you know, the effects should be lower than for other advanced countries. OK, so either there is less demand or there are maybe a Lower willingness to pay for a higher price for this more environmentally friendly products in And more developing countries. So maybe I can sum up this paper. So what I found is environmental certification leads to an increase in the export value by around 10%, which is explained by a price premium of a bit more than 2% and an increase in quantity exported by 9.6%. So environmental certification seems to act as a quality signal for um for these products, for these firms. Uh, but it also creates new economic opportunities for certified firms. So the probability to exports increased by 1.1% following the certification. And the results are heterogeneous across destinations, products, and firms. Thank you very much. Thank you, Camille. Uh, let me open to the questions. Please feel free to ask. Any questions? Marilla. Hi Camille, thank you so much. It's just, I mean, it's amazing work, very, very fascinating, lots of interesting things, uh, that we've seen, and I just have a couple of uh questions, like going back to your first presentation. I was really interested in this regulatory dissimilarity index, and I was, it seemed to me if I paid enough attention that it stopped in 2019. I was wondering if it could be expanded and why did you stop in 2019. Does, does anything change in the. Alright, so just curiosity for that, uh, now second, you said, um, that you showed us the differences in certification impacting exports to developed and developing countries. Uh, yes, the coefficient for developing countries was a little bit smaller, but, um, they kind of look pretty much the same to me, so I'm not sure if we can really speak about heterogeneity. And another, but again, um, you can clarify if I didn't pay attention to the, to the, the numbers, uh, um, enough, but, um, and then the final thing you just said that the probability of exporting increases by 1%. I was a bit kind of disappointing. I would disappointed. I would have expected a bigger result assuming that it's a costly process to certify and, and so on. And regardless of what's driven, what's driving what, you know, the, the willingness to exports driving the need to certification or, or the other way around, it seemed like a very small effect, uh, for a supposedly significant effort. I was wondering what's, what's your take on this, this probability of 1.4% increase in exports seemed, seemed low to me. Uh, thanks so much. Um, so thank you very much. On your first, um, question, um, we just stopped in 2019 cause that's when we started to Um, when we started collecting the data, but we could easily update it and then add the new ones, um. It's just when that's when the all the data stops, so I guess that's why, um, but that could easily be extended, um. On your second point, um, you're totally right. The It's more heterogeneous with the least developed economies. So the least developed exports to the least developed economies, the results were negative. Um, but indeed, it was, um, positive for developing and developed. Um, so in between developing and developed, I'm not, um, I could check whether the coefficients are different, but I'm not sure. So, I think, um, you're right. So, it was more, um, With respect to least developed economies, that there was an um heterogeneity. And um on your last point, um, so I agree, I think, um, So I was also expecting a bigger results, um. Um, so one explanation I might have, um, is the It might be because French firms are already exports to Quite a lot of um destination already even before getting certified. So that might be one explanation. Um, because I think the last time I found some paper um that we're looking, I think in Colombia and Argentina and there or Ecuador and Argentina and then they were mostly uh finding an effect on the extensive margin. So then maybe going back to what Chavi was mentioning, maybe for more developing countries, um, The effect could be even bigger for firms and um allow firms to export even more to more destinations and to reach more markets. So I think maybe one reason why the magnitude is quite small is because France is already quite developed economies and um maybe firms are already exporting to quite a lot of markets. Thank you. That makes a lot of sense. Sorry, just to follow up, Chavi, and I, I promise, um, I was wondering if the reason, uh, you focused, uh, on France, uh, well, it has, uh, something to do with the data quality as well. I'm sure you have research interests in France, but do you know if similar data exists for developing countries that would allow someone that would allow you to replicate this work for developing countries, or? Oh, it's really like, you know, the, because I know that French data is, is super good, but are, I don't know about the quality of this specific information for developing countries, would that be replicable for in other contexts? Um, so the main reason was data availability. Um, it's quite nice to have. Um, so I had a meeting with the International Standard Organization, which are also really interesting in knowing more what's the effect for more developing economies. Um, So far, I think it's quite hard to find information at the firm level on whether a firm is certified or not. Um, So I was looking, um, so I don't know if the accreditation bodies um in the countries would be willing to share the information on which firms are certified or not. Um, this, I don't know. But indeed, I think that would be really interesting. Um, so I think it must exist for Argentina and Ecuador because I saw it for Um, but I think Argentina was another standards, but I think might, um, exist for other countries. Um, but I guess the The aggregate, um, yeah, maybe some firms are willing to share because otherwise it would, uh, be, I think, a long process to check online whether a firm is certified or not, um, if you have to do it one by one. Thank you. Thank you. Any more questions? No, Perhaps I have a, a final question, at least on my end, um, on the terogeneity across sectors. Uh, you look a little bit for the first paper on intermediates, but, uh, I was wondering for the, for the second paper, probably environmental standards. Matter differently for different sectors and exports. I, I was wondering if you, if you explore the differences across sectors of, of French exporters. Um, so I start looking at, um, I wanted to know, for instance, if the sector is more concentrated, maybe if the effect is different, um, So I started looking into this, um, Maybe I could check a bit more, look into um whether Depending on sector characteristic if the effect is different, that's a good point. I haven't checked it yet. OK, colleagues, uh, if there are no more questions, um, I think we, we can close here. Uh, thank you very much, uh, Camille, for a very inspiring discussion, and, uh, the next seminar will be in May 15th. The standard sizes the government role in diffusion of mass production techniques in the US. So please stay tuned. Thank you. Thanks everybody. Thank you very much. Thank you for having me.
showAllTranscripts
no
duration
PT1H10M40S
scene7File
worldbank/WDR2025_Navigating_Standards
scene7Domain
https://worldbank.scene7.com/
scene7FileAvs
worldbank/WDR2025_Navigating_Standards-AVS
title
WDR2025 Navigating Standards
description
WDR2025 Navigating Standards
showTimestampAndTranscript
yes
col-xs-12
col-sm-12
col-md-3
col-lg-3
col-xs-12
col-sm-12
col-md-7
col-lg-7
  • add-style
  • lp-body-content
In this World Development Report 2025 seminar series, Camille Reverdy (Post-Doctoral Researcher, Geneva Graduate Institute) discusses "Navigating Standards: The Impact of Technical Regulations and Voluntary Certifications on Firms’ Trade Strategies" with chair Xavier Cirera, Senior Economist, World Bank.
lp-heading-top-medium
lp-heading-bottom-medium
col-xs-12
col-sm-12
col-md-2
col-lg-2