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:f196b232-6523-41ec-a3f7-28ae42ad4922/play?assetname=KCP_III_private_sector.mp4
keyFrameImage
videoDescription
The Knowledge for Change Program’s 20th-anniversary celebration (“KCP20+”), will be implemented through a series of events. This event focuses on "20 Years of Research and Knowledge on Trade and Private Sector Development."
timestamp

00:06 I'm just going to uh turn directly to our,

00:10 uh,

00:10 two presenters.

00:11 So we're gonna have the same format as the previous um session.

00:13 So we've got,

00:14 uh,

00:14 Uh,

00:15 two of our deck colleagues who are going to kick off the discussion with,

00:18 uh,

00:18 with complimentary presentations,

00:20 and,

00:21 uh,

00:21 and then we'll have 3 panelists,

00:23 and,

00:23 uh,

00:24 I had some inside information that unfortunately one of our

00:26 panelists had a flight canceled and could not join us.

00:29 So I only have 3,

00:30 which buys us the 10 minutes that allows us to wrap up on time.

00:33 So

00:33 hopefully that will work out.

00:35 Um,

00:36 so

00:37 let me,

00:37 uh,

00:37 sorry,

00:37 who's going first,

00:38 Jorge?

00:39 OK,

00:39 so,

00:40 um,

00:40 so we have,

00:41 um,

00:42 Jorge,

00:42 not Jose,

00:43 um,

00:44 showing up,

00:45 uh,

00:46 to,

00:46 uh,

00:46 do the,

00:47 uh,

00:47 to,

00:47 to,

00:48 to kick off the discussion followed by,

00:49 uh,

00:50 uh,

00:50 by David McKenzie.

00:52 Um,

00:52 so

00:53 go ahead and I'll,

00:54 uh,

00:54 warn you around 12 or 13 minutes.

00:56 Uh,

00:57 OK,

00:57 perfect.

00:57 Thanks.

00:58 Thank you,

00:58 Art.

00:59 Uh,

00:59 just as a matter of organization,

01:00 um,

01:01 um,

01:02 let me see,

01:03 um.

01:06 Do

01:06 we have the

01:08 But what this comes from,

01:09 oh yeah,

01:09 OK,

01:10 um,

01:11 uh,

01:12 we,

01:12 the way we organize the presentation in rather than being comprehensive,

01:15 we actually selected a few topics that we think,

01:17 uh,

01:18 uh,

01:18 there's data gaps and data and knowledge gaps,

01:20 and we,

01:21 in the process of presenting these different,

01:23 different gaps,

01:24 we try to highlight the contributions of KCP over the years.

01:27 Uh,

01:28 and I would like to start with a historical perspective of,

01:30 of,

01:30 of,

01:31 of,

01:31 of the,

01:31 uh,

01:31 dimension of the private sector,

01:33 uh,

01:34 development issues

01:35 because,

01:35 uh,

01:36 the 20 years of the KCP,

01:38 uh,

01:38 match very well the development of the PhD agenda.

01:42 The

01:43 KCP started funding considerably,

01:46 uh,

01:47 considerable amounts of,

01:48 of data collection

01:49 efforts on firm level service,

01:51 uh,

01:51 which coupled with the WDR of 2005 on,

01:54 on 2005 on investment climate

01:56 and of course previous resource at the bank on.

01:59 The binding contrast to private sector growth

02:01 explain the generation of the creation of the two main

02:04 diagnostics of the private sector,

02:06 uh,

02:07 that have been,

02:08 um,

02:08 the drivers in a sense of most of

02:10 the operational policy in private sector development,

02:12 enterprise service,

02:13 and the doing business indicators,

02:15 um.

02:17 From here the link between research and

02:20 um

02:21 policy is in the cases of the in the case of the

02:23 private sector development space mostly driven through uh the creation of these

02:28 indicators that as we know are a major tool of our operational

02:31 people to to in uh in here at the bank and our clients

02:35 to galvanize all these principles of research into actual policy.

02:40 Um,

02:40 over the years,

02:41 KCP has funded

02:43 others,

02:43 the,

02:44 the research that,

02:44 uh,

02:45 substantiates other indicators like Finex,

02:47 global Financial development in,

02:48 in,

02:49 uh,

02:49 uh,

02:50 development database,

02:51 and for example,

02:52 this issue,

02:52 the,

02:53 the,

02:53 the,

02:53 the childcare development or childcare services

02:56 for the,

02:56 uh,

02:56 the.

02:57 Important childcare development services have been

02:59 demonstrated first in research and then had been made the

03:02 case for incorporation into women's terms of the law.

03:04 This is a very specific example of the

03:06 of how the I see the

03:08 the process of the linkage between research and

03:11 a policy action has taken place in in PSD.

03:14 Um,

03:15 moving on to the,

03:17 to highlight some of the issues that we wanted to highlight,

03:20 uh,

03:20 in,

03:20 in the issue of informality,

03:22 KCP has funding similar research,

03:25 uh,

03:25 on,

03:26 on,

03:26 on,

03:26 on the,

03:27 on the sector,

03:28 and,

03:28 and,

03:28 and that is,

03:29 uh,

03:30 for example,

03:30 I'm gonna draw again on the work of David

03:32 McKenzie and Christopher Guro that are the panelists today

03:35 on Sri Lanka in 2008 2010.

03:37 9

03:38 in this quite a few

03:40 papers and the research papers that have produced by them,

03:42 but in here I want to highlight the,

03:44 the fact that

03:45 uh this,

03:46 this was,

03:46 is a seminal because it,

03:47 it funded the creation of all the,

03:50 the,

03:50 the whole generation of data

03:52 from the,

03:53 from the scratch,

03:53 from starting from the full enumeration of all

03:56 the informal businesses that operated in Sri Lanka,

03:58 even in the cities of Sri Lanka where the study took place.

04:01 Um,

04:01 this is important because I want,

04:03 what,

04:03 uh,

04:03 I wanted to highlight that because really,

04:05 this really addressed

04:06 one,

04:06 the first,

04:07 uh,

04:07 data gap that I wanted to highlight on an informality,

04:09 which is the lack of census data,

04:12 of any data for that matter on informality in many countries.

04:15 Uh,

04:16 there's very limited information on the,

04:18 on the extensive margin,

04:19 the

04:20 type of the businesses that operate informally and also in the intensive margin,

04:24 uh,

04:24 businesses that operate,

04:25 uh,

04:26 um,

04:27 with informal practices or formal business operate with informal practices.

04:31 Um,

04:32 we have been working on the former,

04:34 uh,

04:34 at least in our,

04:35 in our unit,

04:36 with,

04:37 with that incorporating,

04:38 um,

04:38 new techniques,

04:40 geocoded data collection,

04:41 coupled with

04:42 a new sampling strategy

04:44 that,

04:44 uh,

04:45 takes advantage of the clustering nature of these businesses in,

04:47 in,

04:48 in,

04:48 in,

04:48 in,

04:48 in,

04:49 in,

04:49 in,

04:50 in general in the world.

04:51 And um

04:52 by um

04:54 this method,

04:54 we are able to generate representative samples of the sector in a given region.

04:59 That allow us to study whatever it was,

05:01 you know,

05:01 to do our,

05:02 our,

05:02 our,

05:03 our,

05:03 a questionnaire,

05:04 to implement our questionnaire and at the same time,

05:06 obtain an estimate

05:08 of the population totals,

05:10 in other words,

05:11 the size of the sector.

05:12 Uh,

05:13 this is something that I'm here I'm showing,

05:15 for example,

05:15 as you can see in this,

05:16 uh,

05:17 for this is from the paper

05:18 that we have written.

05:19 Uh,

05:19 we

05:20 got hold of administrative data of census data for Ezotini.

05:24 Uh,

05:24 uh,

05:24 that was fortunate because that allows us

05:26 to actually evaluate the relative merits of

05:29 this methodology vis a vis simple random sampling or certified random sampling,

05:33 and in this paper,

05:34 we show

05:34 that

05:35 there are efficiency gains in terms of the level of effort required to,

05:40 and to implement these studies

05:41 and not only that,

05:42 we can,

05:43 we're able to implement this in 3,

05:45 in between 2 and 4 months in a,

05:46 in a given region.

05:47 Um,

05:49 Moving on to more subject matters,

05:51 uh,

05:52 and to actually

05:53 highlight a few of the knowledge gaps

05:55 in informality as well

05:57 that we find,

05:57 I want to highlight three points.

05:59 The first one,

06:00 again,

06:00 and I'm,

06:00 I'm drawing from the same work of David McKenzie,

06:03 this is,

06:03 uh,

06:03 in fact,

06:04 this slide is

06:05 David's,

06:06 uh,

06:06 but in this one is the,

06:07 the lack of evidence of the benefits of formalization and the reduction of,

06:11 of cost of registration.

06:13 Uh,

06:13 this has been shown by them and many of,

06:15 uh,

06:15 several other,

06:16 other,

06:17 other studies in here.

06:18 This is again with the Sri Lanka study,

06:20 as you can see on the left.

06:21 There's a sharply decreasing demand for formalization

06:24 and this is backed out for an experiment in which,

06:27 er,

06:28 informal entrepreneurs.

06:29 were offered free registration in randomly assigned monthly profits.

06:35 And as you can see on the right-hand side,

06:38 the cumulative distribution of profits

06:40 of the control and the treatment is basically,

06:43 pretty much the same.

06:44 They look alike.

06:45 It is interesting that in this result,

06:47 that uh the lack of evidence of uh

06:50 uh

06:50 uh benefits of formalization in 2008,

06:52 2009 in Sri Lanka,

06:55 the authors mentioned that,

06:56 uh,

06:57 uh,

06:57 by interviewing these,

06:58 these entrepreneurs,

06:59 most of them expressed that they actually were aware

07:02 of the,

07:02 the process of formalization.

07:04 They just wouldn't want to know where are they taking it.

07:06 We,

07:06 I would like to compare this to our recent work in Africa,

07:09 using that methodology that I just mentioned that allow

07:11 us to do a representative sample of the sectors

07:14 in several countries.

07:15 In the,

07:15 in the context of Africa,

07:16 most of these entrepreneurs actually do respond that

07:19 they do not know anything much about formalization.

07:21 It's something that is not in the,

07:22 in the greater skin as an option.

07:25 Um,

07:26 two other points on the data and knowledge gaps that I would like to highlight is the,

07:31 the one on the left is something that has been shown by many studies already,

07:34 uh,

07:35 which is the,

07:36 uh,

07:36 the lower levels of productivity of those

07:38 informal businesses when compared to similar microfirms.

07:42 As you can see here,

07:43 these results come from a paper written so,

07:45 with the,

07:46 with the data that we have recently collected.

07:48 Uh,

07:48 and,

07:49 uh,

07:49 the informal firms are,

07:50 uh,

07:50 uh,

07:51 as you can see in,

07:52 in,

07:52 in,

07:52 in blue,

07:53 and the,

07:53 the micro firms are firms that are very much like the informal firms

07:57 in terms of sector size and even the business environment they face.

08:00 Uh,

08:01 the only differences will be fine between them

08:03 is in the management practices,

08:06 education of the entrepreneur,

08:07 and the,

08:07 uh,

08:07 uh,

08:08 experience.

08:09 Uh,

08:09 we think we still have to do a little bit more digging,

08:11 more research

08:12 to see these are the,

08:13 the real source of such differences in productivity of the rest of,

08:17 are there other things at play.

08:19 A third point as well on the,

08:22 on the knowledge gaps that I would like to highlight

08:24 in the,

08:25 in,

08:25 in the,

08:25 in,

08:26 is the effect of the presence of informality on the formal sector.

08:30 This is,

08:31 uh,

08:31 on the right-hand side,

08:32 the panel comes from a paper also produced here from Amin and others,

08:36 uh,

08:36 in which,

08:37 with more than 100 countries around the world,

08:40 and as you can see,

08:41 there's a clear difference between the firms

08:42 that face competition from formal firms,

08:45 the frames competition for informal firms

08:47 for,

08:48 and those ones that do not face competition for informal firms.

08:51 This result is a result that other

08:53 Papers,

08:54 uh,

08:54 other,

08:54 other authors are also found with all the different uh metrics,

08:58 uh,

08:59 innovation,

09:00 TFP in a different direction.

09:01 So,

09:01 I don't wanna make too much about the difference here in labor productivity,

09:05 but more of the mechanisms at play,

09:08 er,

09:08 that we need to research a little bit more because on the one hand,

09:12 uh,

09:12 one would think that,

09:13 uh,

09:13 the presence of informality is more competition for formal.

09:17 Firms

09:17 and logic would say this firms will become more effective and more innovate more,

09:22 and more productive.

09:23 Uh,

09:23 but on the other hand,

09:24 this access,

09:25 these informal firms have access,

09:27 uh,

09:27 that is in a sense,

09:28 preferential or unequal,

09:30 unfair competition,

09:31 and

09:32 therefore,

09:32 they,

09:33 they have access to,

09:34 uh,

09:35 inputs and

09:36 services,

09:36 uh,

09:37 at lower cost,

09:38 sometimes even free,

09:39 illegally.

09:41 Um,

09:42 let me just move back

09:44 quickly to the formal,

09:45 to the formal sector because I,

09:47 I don't really want to give the impression that

09:48 by all the diagnostics on the first slide and,

09:50 and the,

09:51 and the,

09:51 all the operational work that has derived from that,

09:53 that the,

09:53 in the formal sector,

09:54 there are no

09:56 data gaps.

09:56 And I think one,

09:57 in the presenting here,

09:58 one,

09:58 and I'm less optimistic that we will be doing,

10:00 we will be able to do

10:02 uh something.

10:03 effectively in the medium term,

10:05 which is the lack of administrative data

10:07 to produce research.

10:08 And I'm using here

10:09 the a KCP funded project from the Mary Howard and

10:12 others did in 2013 in which they try to collect

10:16 uh

10:16 these indicators on firm dynamics,

10:18 productivity,

10:19 and,

10:20 and,

10:20 and,

10:20 and resource allocation

10:22 uh from administrative data.

10:24 10 years later,

10:25 now 223,

10:26 it's only

10:27 a handful of countries where we have administrative data

10:29 and the issues,

10:30 uh,

10:30 the obstacles that we face in accessing,

10:32 uh,

10:33 administrative data,

10:34 uh,

10:34 are,

10:35 are relatively serious in my view because there are

10:37 legal issues on how the data was collected,

10:39 uh,

10:39 and,

10:40 and difficult to overcome.

10:41 And when one overcome those,

10:43 those issues,

10:44 these issues of comparability and quality.

10:45 So,

10:45 I think this is an issue that will remain

10:47 er for,

10:48 for some time.

10:48 And if you follow my premise here,

10:51 I think the

10:53 usefulness of the relevance of survey data becomes even more relevant,

10:56 more important.

10:57 Uh,

10:57 and so I want to highlight two points I saw as well that we,

11:00 we think there is some,

11:02 uh,

11:02 knowledge gaps and data gaps in survey data.

11:05 One is related to a point that was being discussed earlier,

11:07 is the need to have more data,

11:09 uh,

11:09 tailored for services and to understand,

11:11 well,

11:12 to,

11:12 to understand better multi-factor productivity for manufacturing as well.

11:15 But mostly for services

11:17 and the second point that I would like to highlight is the need to have more um

11:22 frequent and

11:24 uh uh frequent and uh a

11:26 larger panel datasets

11:27 that will let us address the issues,

11:29 the issues of causation and move away from the issues of correlations.

11:33 Uh,

11:34 I don't wanna over,

11:35 overstress this because,

11:37 uh,

11:37 KCP has funded quite a bit of research,

11:39 uh,

11:39 that,

11:40 uh,

11:40 on this and a lot of cross-section data

11:42 that has been very illuminating,

11:45 uh,

11:45 on the evaluating,

11:46 for example,

11:47 the efficiency of the state-provided services for the private sector.

11:50 Here,

11:50 I have one paper that I find very,

11:52 very.

11:52 Relevant,

11:53 which is by,

11:54 by Caroline Freund and Bod Rikers and others here in DC,

11:57 in which they have shown that,

11:58 uh,

11:59 and tested

12:00 that the hypothesis of greasing the wheel of bribery,

12:02 that it is,

12:03 it,

12:03 that it,

12:04 that it does not hold.

12:05 As you can see in this example here is one of the

12:07 many examples of different transactions that firms interact with the government.

12:12 Um,

12:13 let me just

12:14 finish here,

12:15 uh,

12:15 uh,

12:16 that with one final point that I would like to,

12:18 uh,

12:18 highlight,

12:18 which I think is an area in which we

12:21 will need in thinking on what I was saying

12:23 earlier on thinking on to the future,

12:26 uh,

12:26 which we,

12:27 we,

12:27 and I think this area is one because we need,

12:29 it's interesting and relevant.

12:30 Because not only is it important,

12:32 but we need to think not only on

12:33 methodologies,

12:34 how to address it,

12:35 but also

12:35 new data that we need to be collected.

12:37 And,

12:37 and this is the,

12:38 the economic cause of discrimination

12:41 along the lines of gender,

12:42 race,

12:43 and even sexual identity.

12:44 These are populations that are difficult to,

12:46 to,

12:47 to,

12:47 to collect data from,

12:49 uh,

12:49 and,

12:49 uh,

12:49 and there is not quite clear even how to go about measuring this economic cost.

12:53 That's what I'm using here.

12:55 Uh,

12:55 this very relevant paper and I find very impressive,

12:58 uh,

12:58 from in Econometrica in 2019,

13:01 which does this,

13:02 at least for the US and as you can see,

13:04 it compares the GDP growth

13:06 of the,

13:06 the actual GDP growth in blue

13:08 with the GDP growth hypothetical that will emerge from their model,

13:12 uh,

13:12 if the restrictions,

13:14 uh,

13:14 the discriminatory practices of black men and women in

13:17 the US in the 1960s will have prevailed.

13:19 I can,

13:20 you can see that it's clear in the graph,

13:21 the,

13:21 the big difference between the two,

13:23 which in the paper they say is about,

13:25 uh,

13:25 accounts for about 40% of the GDP growth uh between 1960 and 2010,

13:30 which I think is,

13:31 it makes the case for how important this is.

13:33 And with that,

13:34 They did

13:38 Great,

13:38 thanks.

13:40 So,

13:41 um,

13:42 I'm gonna dig in on sort of one topic that the KCP has helped illustrate this sort of

13:49 going back and forth between data,

13:51 research,

13:51 policy,

13:52 and,

13:52 and back to more data and more,

13:54 more research.

13:54 And so,

13:55 uh,

13:56 this is the sort of,

13:58 Question of,

13:59 you know,

13:59 why the firm size distribution looks the way it does in most developing countries.

14:03 And so,

14:04 you know,

14:04 if we look,

14:05 most firms in developing countries are very small,

14:08 many are informal,

14:09 their productivity is very low,

14:11 uh,

14:11 and so,

14:12 you know,

14:12 this raises

14:13 three questions about,

14:15 you know,

14:15 why are there so many small and unproductive firms?

14:17 How much of a problem is,

14:19 is this,

14:19 uh,

14:20 and,

14:20 and their informality,

14:21 and then,

14:21 you know,

14:22 what can policy do about it?

14:24 And so,

14:25 you know,

14:25 if we want a framework

14:27 for thinking about,

14:28 you know,

14:28 the firm's size,

14:29 then we can sort of think about

14:31 the problem of a typical firm owner who's

14:33 trying to choose their inputs of capital and labor

14:36 to maximize their profits,

14:37 and they have some sort of production function,

14:39 which has got this,

14:41 um,

14:41 capital and labor that they're buying,

14:43 and then this term theta that they're,

14:45 um,

14:46 using that,

14:46 that helps them turn their capital and labor into,

14:49 to output.

14:50 And so this suggests,

14:51 you know,

14:51 3

14:52 potential explanations for firm size,

14:54 um,

14:54 being the way it is.

14:55 So one is that firms are small because they're constrained,

14:57 and we have,

14:58 you know,

14:59 a whole other programs on finance that we're,

15:01 we're not gonna talk about today or on labor

15:03 markets that we're not gonna talk about today,

15:04 but thinking about

15:05 sort of constraints for firms and getting the inputs that they want.

15:09 The second that I'm gonna really dig into is that firms are small because they have,

15:13 uh,

15:14 a low theta.

15:14 They're just not very good at,

15:16 um,

15:16 turning their inputs into output,

15:18 and so we're gonna dig into that.

15:19 And then the,

15:20 the third point,

15:21 which is,

15:21 um,

15:22 sort of related to the informality point that Jorge was talking about is that,

15:25 you know,

15:25 firms are small because the regulations are affecting their choice of,

15:28 you know,

15:28 their production function here,

15:30 should they be formal or informal,

15:32 and,

15:32 and that can also affect the cost of their inputs.

15:35 But I'm gonna dig into this sort of middle explanation,

15:38 which is,

15:39 Um,

15:39 you know,

15:39 maybe firms are so small,

15:41 because they're just not very good at,

15:42 at managing their,

15:43 um,

15:44 process of converting inputs into outputs.

15:47 And so,

15:48 uh,

15:48 you know,

15:49 how do we go about measuring and improving

15:51 this,

15:52 this,

15:52 uh,

15:52 term here?

15:53 Well,

15:54 you know,

15:54 where were we 20 years ago?

15:56 Sort of 20 years ago,

15:57 there was this recognition that there

15:59 was something that we called managerial technology

16:02 that affected the ability of firms to translate these inputs into outputs.

16:05 It's,

16:05 but it was really a black box,

16:07 um,

16:07 so,

16:08 uh,

16:08 you know,

16:09 there's famous papers that sort of say,

16:10 whatever this is,

16:11 it's,

16:11 you know,

16:12 some firms are better at it than others,

16:13 um,

16:14 or there was this sort of view that management was really,

16:17 you know,

16:17 uh,

16:17 to do with style,

16:19 or with charisma,

16:20 or,

16:20 you know,

16:20 something that was,

16:22 was hard to

16:23 potentially teach,

16:24 and some people were just born with it,

16:25 and others,

16:26 um,

16:26 were,

16:27 were not.

16:28 Now,

16:28 you know,

16:28 of course,

16:29 there was a,

16:29 a wide range of business training programs that were based on

16:32 the idea that there were some things that you could teach,

16:34 um,

16:35 but there weren't really a lot of evaluations or no evaluations of these programs,

16:38 and there was less policy attention

16:40 to what could be done in larger firms to improve this as well.

16:44 So,

16:45 uh,

16:45 this is where the sort of first step of,

16:47 of really just trying to measure and understand,

16:50 um,

16:50 comes through.

16:51 And so,

16:52 uh,

16:52 you know,

16:52 there was a set of influential,

16:54 um,

16:56 surveys done in large firms by Nick Bloom and John Van Reenen

16:59 that,

17:00 you know,

17:00 really shaped the idea that these,

17:03 That management was,

17:04 was not just a,

17:05 a style or something that you're born with,

17:07 but it was really a set of concrete practices about how do you manage good workers

17:11 and bad workers,

17:12 how do you deal with quality,

17:13 how do you plan for the future,

17:15 things like that.

17:16 And in the KCB supported work,

17:18 uh,

17:19 Chris Woodruff,

17:19 who's,

17:19 who's gonna be one of the panelists,

17:20 and I

17:21 thought about how do we,

17:23 Take those same ideas and,

17:24 and measure this in small firms,

17:26 and do,

17:26 do these things matter in small firms?

17:28 If you walk into a lot of small firms,

17:30 they're not keeping records,

17:31 they're not doing any marketing,

17:32 they,

17:33 they don't have a budget,

17:34 does this matter?

17:35 And so what we see is,

17:36 you know,

17:37 across a range of countries,

17:38 both in the cross section over time,

17:40 it seems like,

17:42 Firms differ a lot in these practices,

17:44 that firms that do more of these practices

17:47 are more profitable,

17:48 sell more,

17:49 and um grow faster.

17:51 And so it seems like there's something about,

17:53 you know,

17:53 business practices that would be good to try and get firms to,

17:56 to improve.

17:58 Uh,

17:59 this is also something that the,

18:01 um,

18:01 enterprise surveys unit has been building on with,

18:03 with,

18:04 um,

18:04 adopting,

18:05 how do you take these,

18:06 um,

18:08 ways of measuring management that were done,

18:10 sort of,

18:11 uh,

18:12 um,

18:12 in this sort of very expensive double-blind

18:15 system and adopt them

18:17 to how they can be done in firm surveys.

18:19 And so,

18:20 you know,

18:20 they've been adopting both of these large and small firm,

18:23 um,

18:24 techniques to now make this available for a lot more countries.

18:27 So then,

18:28 you know,

18:28 once we have

18:29 this idea that,

18:30 that there's something about these practices that,

18:32 that seem to matter,

18:33 the question is,

18:34 you know,

18:35 can we do something to improve them,

18:37 and,

18:37 you know,

18:38 does this,

18:38 this matter?

18:39 And so this is where I wanna sort of

18:40 give an illustration of this interplay between,

18:43 You know,

18:43 the KCP supporting

18:44 what I call blue skies research,

18:46 where it's not really testing a policy that we think about

18:49 testing,

18:50 it's not linked closely necessarily to a

18:52 particular country operation that we have at the moment,

18:55 but it's trying to generate,

18:56 you know,

18:56 where should be things,

18:57 things be going in the future.

18:59 Um,

19:00 you know,

19:00 is,

19:00 is there even a,

19:01 should we even be trying to push policy in this direction?

19:04 And so the KCP supported this original work that,

19:08 uh,

19:08 I did

19:09 in India with,

19:10 with Nick Bloom and,

19:11 and others on

19:12 sort of a proof of concept,

19:13 which was,

19:14 if you take a badly managed firm,

19:16 can you actually go and improve management in those firms,

19:18 and would it actually matter?

19:20 And so,

19:20 you know,

19:20 this was

19:21 definitely not something that we'd recommend for policy.

19:24 We hired the most expensive consultants you could find on the planet,

19:27 probably.

19:28 Um,

19:29 uh,

19:29 you know,

19:29 very expensive global consulting firm

19:31 came in,

19:32 did

19:33 really intensive consulting on these firms,

19:36 uh,

19:36 and,

19:37 and,

19:37 you know,

19:37 spent

19:38 months in these firms

19:39 and improved management practices.

19:41 But the key thing here was that we could see,

19:43 once you improve management practices,

19:45 we could track in these data that you could get quality improvements

19:49 and see total factor productivity improve.

19:51 And so,

19:52 We could,

19:53 you know,

19:53 show not just that there was this

19:54 association between better management and better practices,

19:57 but that there was a causal link.

19:59 If you could improve management,

20:01 then

20:01 you could get this,

20:02 this increase in productivity.

20:04 So this was,

20:05 you know,

20:05 nice,

20:06 but this was not,

20:07 you know,

20:07 where

20:08 you would necessarily want to go with,

20:09 with policy,

20:11 doing the exact same thing.

20:12 And so,

20:13 you know,

20:14 then,

20:14 um,

20:15 but this is how this sort of research can then help generate

20:18 new operational ideas.

20:19 And so my operational colleagues,

20:21 you know,

20:21 are talking with finance ministers,

20:24 productivity is always mentioned as an issue.

20:26 And the question is,

20:26 you know,

20:27 what have you got that,

20:28 that tells us

20:29 how we could improve productivity?

20:30 And they said,

20:30 well,

20:31 you know,

20:31 there's this work from India that shows

20:33 that management can,

20:34 can help.

20:35 Um,

20:35 and,

20:35 and,

20:36 you know,

20:36 is there a way We could think about doing that in Colombia.

20:38 And so then I worked with them in,

20:40 in,

20:40 in the government in Colombia,

20:41 and we sort of tested,

20:43 could you do this,

20:43 but could you do this in a more scalable,

20:46 more cost-effective way?

20:47 And we tested a group-based consulting,

20:49 uh,

20:50 method,

20:50 as well as this individual consulting method.

20:52 And you see in this graph,

20:53 you know,

20:54 across a whole range of management practices,

20:56 you get improvements,

20:57 um,

20:57 that are very similar from this group-based consulting approach than you.

21:01 from,

21:01 from the individual approach at a third of the cost.

21:03 So,

21:04 you know,

21:04 we get this,

21:04 this method that can improve,

21:07 um,

21:07 some growth in Colombia,

21:09 can be done at a lower cost,

21:10 and,

21:11 you know,

21:11 then that link,

21:12 links to,

21:13 sort of,

21:13 the next round of policy where now,

21:15 now that that's been done in Colombia,

21:17 countries are coming and saying,

21:19 oh yeah,

21:19 let's see if we can do this in Ghana,

21:21 we can do this in Malawi,

21:22 etc.

21:23 And so,

21:24 um,

21:25 You know,

21:26 that's,

21:26 that's sort of,

21:27 you know,

21:27 from that,

21:28 that blue skies research leading to new government projects.

21:32 And then,

21:32 you know,

21:32 one of the things that comes up is that there's just not many of these firms.

21:36 We're trying to do this a lot with sort of small samples.

21:38 And so,

21:38 one of the things that the KCP has also been,

21:41 um,

21:41 helping us do is more methodological improvements on,

21:44 you know,

21:44 how do we incorporate prior information,

21:46 how do we make the most of these small samples.

21:48 And so some of the

21:49 most recent work we've been doing on

21:51 uh new Bayesian impact evaluation methods has been supported by the KCP there.

21:56 So then,

21:57 you know,

21:57 that's large firms.

21:58 What about small firms?

21:59 Well,

22:00 you know,

22:00 some of,

22:00 uh,

22:01 one of our panelists,

22:02 Dean Kalin was,

22:03 you know,

22:03 really one of the first,

22:04 I think the first to,

22:05 to look at,

22:06 um,

22:07 was this business training effective.

22:09 And there's been sort of many studies since that have found that,

22:12 you know,

22:12 some of these business training programs

22:14 can help

22:14 firms,

22:15 even the very smallest firms,

22:17 improve their business practices,

22:18 improve profits and sales.

22:20 But,

22:20 um,

22:21 you know,

22:21 it's,

22:21 it's

22:22 somewhat mixed evidence.

22:24 Uh,

22:24 in some cases,

22:25 it's not always as

22:26 factors as,

22:27 as we'd like.

22:28 And it seems quite hard to scale some of these in-person classroom-based,

22:32 um,

22:32 training.

22:33 And so,

22:33 that's where,

22:34 you know,

22:35 I think two of the other directions that have been

22:36 coming through some of the most recent KCP supported work,

22:39 I've been doing,

22:39 my colleagues,

22:40 Miriam and,

22:40 and Kayle here have been working in Brazil to see if

22:43 you could sort of start with a much lighter touch approach,

22:46 give some firms diagnostics,

22:48 um,

22:49 use this to identify,

22:50 you know,

22:50 which firms you might then want to support in the second

22:53 stage and get firms to demand some of the training themselves.

22:56 Um,

22:57 unfortunately,

22:58 our,

22:58 our panelist,

22:59 um,

22:59 Geo,

23:00 who,

23:00 who's,

23:00 uh,

23:01 on a flight right now,

23:02 uh,

23:03 she heads

23:03 this,

23:04 uh,

23:05 organization of,

23:05 uh,

23:06 female entrepreneur,

23:07 um,

23:08 training.

23:08 Programs in,

23:09 in Mexico called CREEA,

23:11 and,

23:11 you know,

23:11 we've been working with them to see if you could do,

23:13 use digital technology to try and scale um training and,

23:17 and,

23:17 you know,

23:17 provide it remotely,

23:18 and we've been working,

23:19 you know,

23:19 both in Mexico and Guatemala,

23:21 you know,

23:21 using Zoom,

23:22 use,

23:23 using what we're doing right now to do training to,

23:25 to people all across the country.

23:27 So,

23:27 you know,

23:27 we're,

23:28 we're trying to learn that.

23:29 So,

23:29 just,

23:30 to sort of,

23:30 you know,

23:30 where,

23:31 where do we go next?

23:32 I guess,

23:32 you know,

23:33 there's a couple of recurring themes of persistent puzzles.

23:35 And so,

23:36 one is just,

23:37 you know,

23:37 firms are just very different from one another.

23:39 Management capacity varies dramatically across,

23:42 um,

23:42 and within countries.

23:44 There's just lots of variation in the scope for growth.

23:46 And so the question is,

23:47 you know,

23:47 how can we identify which firms we should be targeting,

23:50 um,

23:50 policies to,

23:51 um,

23:52 more?

23:52 How can we scale this in a way that,

23:55 um,

23:56 You know,

23:56 the market works better,

23:58 how much does the government need to be involved

23:59 versus just trying to get the market to work,

24:02 um,

24:02 better,

24:02 and then we need to,

24:03 you know,

24:03 work on collecting better,

24:04 um,

24:05 data on firm,

24:06 firm outcomes,

24:07 which is quite tricky.

24:08 And so,

24:09 you know,

24:09 as a,

24:10 uh,

24:10 you know,

24:10 then there's these new areas of private sector development that

24:13 are really coming to the fore in terms of,

24:15 you know,

24:15 how do we collect more information on,

24:17 on green growth,

24:18 green PSD,

24:20 um,

24:20 transactional government,

24:21 the enterprise surveys,

24:22 you know,

24:22 have been starting to measure these practices.

24:25 And,

24:25 you know,

24:25 hopefully,

24:25 this is the next phase of,

24:27 you know,

24:27 we collect some data,

24:28 we identify some diagnostics,

24:30 and now we can,

24:31 you know,

24:31 start some interventions based on,

24:33 on that.

24:33 And so,

24:34 um,

24:35 you,

24:35 you know,

24:35 there's,

24:36 there's just sort of a lot of scope now with big data,

24:38 with AI with other approaches to try and improve our,

24:41 our survey technology,

24:42 and I hope this is sort of gonna be,

24:44 you know,

24:44 we start a whole new batch of,

24:46 uh,

24:46 diagnostics that can help get this cycle going again.

24:49 So let me stop there.

24:53 Thanks very much,

24:54 Jorge and David,

24:55 um,

24:56 uh,

24:56 for great presentations,

24:57 uh,

24:58 or presentation,

24:59 I guess very integrated,

25:00 um,

25:01 so we'll turn now to our panelists and as they've already mentioned,

25:04 uh,

25:05 unfortunately,

25:06 uh,

25:06 Giovanna Hernandez Constantino,

25:08 um,

25:09 who was supposed to be our first panelist is

25:11 unable to,

25:11 to join us because of,

25:13 uh,

25:13 travel mishaps,

25:14 um.

25:15 Uh,

25:15 but we do have,

25:16 um,

25:17 uh,

25:17 3 other excellent panelists who we'll go to in turn in the same format as before.

25:22 So,

25:22 uh,

25:22 we'll begin with Dean Carlin,

25:24 who,

25:24 uh,

25:25 needs a little introduction,

25:26 um,

25:26 and,

25:27 uh,

25:27 currently is the chief economist at USAID.

25:30 Um,

25:31 then we'll turn to Deborah Revoltella,

25:33 who is the chief economist and director of

25:35 the Economics Department at the European Investment Bank.

25:38 Um,

25:38 and finally,

25:39 we'll turn to,

25:40 uh,

25:40 Chris Woodruff,

25:41 who is at Oxford and,

25:42 uh,

25:43 uh,

25:43 also I think needs a little introduction given the,

25:46 uh,

25:46 the importance of the work that he has,

25:48 uh,

25:48 uh,

25:50 he has done and that we frequently see cited here in the bank.

25:53 Um,

25:54 so,

25:54 starting with you,

25:55 Dean,

25:56 uh,

25:56 we'll,

25:56 uh,

25:57 we'll go with,

25:57 uh,

25:58 my,

25:58 sorry,

25:59 I think the commitment was to 8 minutes each.

26:01 So,

26:01 if you can sort of aim for that and,

26:03 uh,

26:04 Uh,

26:05 and,

26:06 and then we'll have a chance for a discussion.

26:08 Uh,

26:08 so

26:09 over to you,

26:09 Dean.

26:10 OK,

26:10 hold on,

26:10 I'm starting a timer.

26:11 Don't start the 8 minutes yet.

26:14 There we go.

26:15 Um,

26:15 hi,

26:15 everyone.

26:16 I,

26:16 I wish I was over there.

26:17 Um,

26:18 sorry,

26:18 I,

26:18 I ended up too tight,

26:20 um,

26:21 on the,

26:22 on the,

26:22 on the walkover.

26:23 So,

26:24 I'm going,

26:24 I'm doing the lazy path.

26:26 Um,

26:27 so,

26:27 I wanted to share 33 thoughts that,

26:31 um,

26:32 that

26:33 also feed,

26:34 you know,

26:35 build on a lot of the things that we just discussed,

26:37 um,

26:37 and,

26:38 and,

26:39 and are by no means in lieu of,

26:40 I think there's a lot of,

26:42 like,

26:42 really important questions that,

26:44 um,

26:45 that,

26:46 that we just heard about that are being posed.

26:48 There's a lot of different directions where more research is needed to understand.

26:52 You know,

26:52 things like the technology question that David was just talking about,

26:56 um,

26:56 for,

26:57 for achieving scale and,

26:58 um,

26:59 and,

26:59 you know,

26:59 we,

27:00 we have an overall

27:02 I don't want to say conundrum,

27:03 but a little bit of a conundrum of the fact that,

27:06 you know,

27:06 I think,

27:07 I think

27:08 we can kind of go backwards and say,

27:10 if someone asks you for a point estimate,

27:13 You can find the study that,

27:14 that finds it when it comes to things that are

27:16 in the space of promoting managerial capital or business training.

27:20 And,

27:20 um,

27:21 and that shouldn't be seen as a,

27:23 as a,

27:24 in some,

27:24 in some respects,

27:25 I think that should be seen as a motivation,

27:27 not as a,

27:28 not as a deterrent,

27:29 a motivation for,

27:30 for the kind of work that we've just heard about,

27:32 that's really trying to get in more granular understanding of

27:36 Of both,

27:37 how do we do,

27:38 uh,

27:39 diagnostics as well as how do we,

27:41 how do we

27:42 separate out and understand what types of implementation strategies are,

27:46 are,

27:46 are working and the intersection there and hopefully we'll start,

27:49 um,

27:49 getting more and more insights.

27:50 But as,

27:51 as you heard,

27:51 there's,

27:52 you know,

27:52 there's definitely

27:53 You know,

27:54 we,

27:54 we,

27:54 you know,

27:55 the state of the evidence is definitely better than um

27:57 we can,

27:58 that everything goes,

27:59 right?

27:59 It's,

27:59 it,

28:00 we,

28:00 we do,

28:00 there are some clear patterns that are coming through that,

28:03 that show

28:04 um

28:05 policymakers who are not interested in research,

28:07 you just want to use evidence that does show

28:09 we,

28:09 you know,

28:10 we do have important insights as you've just been hearing about.

28:13 Um,

28:14 so I want to highlight three areas that I see as kind of big gaps that in this,

28:19 in this area that,

28:20 and when I say big gaps,

28:21 I want to be clear that like,

28:23 there's definitely like really good,

28:24 smart,

28:25 um,

28:26 careful,

28:26 um,

28:27 policymakers into,

28:28 you know,

28:29 implementers and researchers working on these,

28:31 so,

28:31 you know,

28:32 To a lot of people in the room,

28:33 um,

28:34 there's a

28:35 none of these three are gonna be like,

28:36 wow,

28:36 we never thought of that before.

28:37 But,

28:38 um,

28:38 but I do see these as three areas that are

28:40 continue to be,

28:41 um,

28:43 areas of big,

28:44 big gaps relative to

28:46 the kind of standard,

28:47 um,

28:48 impact evaluation.

28:50 The first is very much motivated by my baldness.

28:53 Um,

28:55 And um,

28:56 although the jokes don't seem to work too well over Zoom,

28:58 but um I have a real gripe with hair,

29:01 um,

29:02 and,

29:02 and with hair salons.

29:05 And it's just to make a very simple point about general equilibrium effects

29:09 that

29:10 we don't,

29:10 you know,

29:11 we talk about them a lot.

29:12 We obviously thought about them a lot.

29:13 There's been a lot of good research on it and,

29:15 you know,

29:16 by,

29:16 by people,

29:17 by David and others,

29:18 and,

29:19 um,

29:19 and,

29:20 but,

29:20 you know,

29:21 obviously,

29:21 no pun intended,

29:22 there's not gonna be a very simple generalized answer to generalized equilibrium,

29:26 general equilibrium effects.

29:28 Um,

29:29 and yet we desperately need to understand more about them.

29:31 Um,

29:32 there's been,

29:33 you know,

29:33 a lot of,

29:34 when we look at the expansion of microenterprise in the US,

29:36 I was always struck by

29:38 how often the prototypical enterprise that,

29:41 that was being promoted as a microenterprise in the United States was hair salons.

29:45 And this is why I make the hair joke,

29:47 because like,

29:48 adding more hair salons in,

29:50 you know,

29:50 in most communities,

29:52 uh,

29:52 you know,

29:52 there's only a fixed amount of hair in the world.

29:55 And,

29:55 um,

29:56 as much as I might resent those people,

29:58 um,

29:58 they're adding more hair salons does not lead to more haircuts.

30:02 Um,

30:03 and so,

30:04 uh,

30:04 you know,

30:05 what,

30:05 what's really happening here,

30:06 even if you set up a nice randomized trial,

30:08 you're very rarely gonna fit,

30:09 you know,

30:09 kind of detect that kind of effect.

30:11 It's,

30:11 you need a very large study,

30:13 um,

30:14 you need a lot of the stars to be aligned.

30:16 But we do need more work to understand this.

30:19 Um,

30:19 and we need to first separate out and not,

30:22 and not do what I just did and talk about general,

30:25 general equilibrium effects as some sort of monolithic thing and,

30:27 and

30:28 separate it out into information effects,

30:31 um,

30:31 and price effects and then kind of industrial organization,

30:34 competition

30:36 structure effects.

30:37 And

30:38 the,

30:38 so that's the first thing we need to do when we think about G and then,

30:41 you know,

30:41 could also think about kind of macro multiplier kind of um effects as well.

30:45 So,

30:46 you know,

30:47 we need to separate out what we mean by this.

30:49 We need to do,

30:49 we need to try to incorporate when we can better,

30:52 better disentangling of these effects.

30:54 And then we,

30:55 you know,

30:56 we need to,

30:56 we need to kind of rope in our macro friends as,

30:59 as micro people and um to,

31:01 to um integrate

31:03 some of the,

31:04 the micro studies into better macro modeling of these kinds of issues to understand

31:08 more about

31:09 how to set policy.

31:11 That um can hopefully use GE effects to improve impacts rather than,

31:16 um,

31:17 rather than,

31:17 you know,

31:18 the opposite,

31:18 which is the fear in many cases that they're going to,

31:21 um,

31:21 diminish treatment effects and,

31:23 and all we're doing is business stealing by promoting some businesses.

31:27 Um,

31:27 the second

31:28 gap that I,

31:30 that I,

31:30 um,

31:31 have

31:32 been,

31:32 um,

31:32 kind of obsessed with for a while,

31:34 and I think it's actually much less,

31:36 much less research compared to like the general equilibrium is,

31:39 um,

31:39 a long time ago I wrote a,

31:40 uh,

31:40 a kind of a thought piece with Sendel Mono,

31:42 and been to Anno called Three Anomalies.

31:46 Um,

31:46 and one of those anomalies was,

31:48 why don't we see more people share

31:50 businesses?

31:52 Two-person,

31:52 three-person equity firms,

31:54 um,

31:54 as the on-ramp.

31:56 And,

31:57 and it's really easy to think through examples and

32:00 household enterprises where there's a clear economy of scale.

32:03 Um,

32:04 by having or economy of scope,

32:06 depending on the business style,

32:08 goat herding,

32:09 piggeries,

32:10 cow sheds,

32:11 where a cow shed is a fixed cost and having multiple households keep their cows.

32:14 There's lots of examples like this that we,

32:17 that are easy to tell.

32:18 And,

32:19 you know,

32:19 what's the obstacle here in understanding more about those,

32:23 um,

32:23 those challenges?

32:24 Is it moral hazard?

32:25 Is it training?

32:26 Is it,

32:27 um,

32:27 is it trust?

32:28 Is it just,

32:29 um,

32:30 is it,

32:30 is it just wrong and the returns aren't that high?

32:32 To some of these,

32:33 but it does seem like on paper,

32:35 some of these returns could be really,

32:37 um,

32:38 really remarkable.

32:39 And,

32:40 and we need to see more work on that.

32:41 So,

32:41 that's not quite on the,

32:42 that's,

32:43 I would put that in the small,

32:44 but not the medium size.

32:45 This is about the,

32:46 that,

32:46 that bridge from getting from micro to small firms.

32:50 And the third

32:52 is

32:53 why,

32:54 here's a puzzle that I've always been struck by is,

32:57 given that we do see

32:59 some

33:00 um really strong impacts from,

33:02 you often more tailored and customized

33:05 managerial capital intervention.

33:07 I'm thinking about things like

33:09 the India study that David did,

33:11 I'm thinking about the Mexico study that I

33:13 did with Antoinette um Shore and Miriam Brune.

33:16 Um,

33:17 and,

33:18 um,

33:19 these showed really high returns.

33:21 Um,

33:22 we saw some pretty impressive improvement in,

33:25 in labor force,

33:26 in returns to,

33:27 um,

33:28 in kind of total factor productivity in,

33:29 in,

33:30 in Mexico.

33:32 So how is it possible,

33:34 and David's heard me ask this question before,

33:36 so,

33:36 how is it possible

33:38 that these consulting firms can be so good.

33:41 At giving out advice,

33:43 and yet they're not.

33:45 Like expanding

33:47 amazingly well and getting really,

33:49 you know,

33:49 rich themselves.

33:50 What,

33:50 what's the,

33:51 what's the problem in the consulting market,

33:54 if this really is such a,

33:56 a clear path to improving SME,

33:59 um,

34:00 total,

34:00 you know,

34:00 profits?

34:01 Where's the issue?

34:02 Is it risk?

34:03 Is there heterogeneity and quality,

34:05 and people don't know how to choose firms?

34:08 Is it information?

34:10 Um,

34:10 is it,

34:11 um,

34:11 is it capital?

34:13 Is it just a capital constraints or a contracting issue?

34:15 And that's my 8 minutes.

34:17 Um,

34:18 you know,

34:18 where's,

34:19 what,

34:19 where's the market failure rate in the market for consulting?

34:22 Like,

34:22 and,

34:23 and I,

34:23 I'm,

34:24 I'm still,

34:25 I,

34:25 I,

34:25 I was struck by this from the study that we did in Mexico.

34:28 I've been struck by this by any,

34:30 any study that shows big success for

34:32 managerial capital interventions from a for-profit entity.

34:36 That is not then expanding magnificently themselves.

34:39 Like what's,

34:40 what's the,

34:41 what's the constraints in that market?

34:44 So I'll close there.

34:47 Thanks very much,

34:48 Dean.

34:49 Um,

34:49 so you want to talk to more macroeconomists,

34:51 I'll just remind you that macroeconomists mean something

34:53 very different when they talk about haircuts.

34:56 Uh,

34:56 so just

34:57 bear that in mind when you start those discussions.

35:00 Um,

35:00 all right,

35:01 over to Deborah.

35:05 Thank,

35:06 thank you very much and thank you very much for inviting me for uh for uh this

35:11 uh panel and uh this uh discussion and I'm also very sorry not to be with you today

35:16 um in Washington but uh being only remotely,

35:20 but for me it would have been a,

35:21 a longer traveling coming there.

35:25 So,

35:25 um,

35:26 I,

35:27 I read the task,

35:29 uh,

35:29 for being in this panel in terms of,

35:31 uh,

35:31 suggesting topics that I think,

35:34 uh,

35:34 in terms of private sector development at this point in time,

35:38 we should,

35:38 uh,

35:39 uh,

35:39 really put in the agenda and also,

35:42 uh,

35:42 think where we need to invest.

35:45 Much more uh on uh data and uh research and I think uh

35:50 um where I'm coming from and uh probably is also for uh the institution I work with

35:57 um one of the uh key concern uh in this moment that uh

36:02 that we are looking at on the research side is trying to understand.

36:07 Um,

36:08 what motivate,

36:09 uh,

36:10 the transition,

36:11 uh,

36:11 the,

36:11 the,

36:11 the transformation of of firms of,

36:14 uh,

36:15 of companies

36:16 in the net zero transition and,

36:19 uh,

36:20 addressing climate change.

36:22 And I think on that point.

36:23 Point of view,

36:24 we really,

36:26 we really have an issue in terms of data availability and understanding.

36:31 We're investing much more in terms of data and the survey also can play a rule

36:37 would be quite relevant.

36:40 Um,

36:41 where I'm coming from,

36:42 uh,

36:43 um,

36:44 we,

36:45 uh,

36:45 we know that,

36:46 uh,

36:46 uh,

36:47 that the transformation will,

36:49 uh,

36:49 will the net zero transition will lead to a complete change in relative prices,

36:55 reallocation of resources,

36:57 and economic power

36:59 within and across sectors,

37:02 and the,

37:02 the incentive.

37:03 to move as a company and to transform really

37:07 change a lot depending on where you position it.

37:11 If you hold

37:13 assets that will become stranded assets,

37:17 if you are in a good position for developing a new technologies or not,

37:23 whether you have access to.

37:26 Resources that are necessary for

37:29 the green transition

37:31 and also what is the policy framework

37:35 around you that incentivize or disincentivize the transition.

37:42 So

37:43 to me what

37:45 what would be a very useful.

37:49 Processes to,

37:51 to start looking more at

37:53 uh what motivates firms to invest in the net zero transition,

37:59 to transform and radically transform

38:02 why some firms

38:03 are lagging behind and others are.

38:07 Accelerating the transition while some

38:11 play into

38:12 gaining the upside while others just

38:15 play into

38:18 amortizing the current assets

38:22 before doing anything in terms of transformation.

38:25 And the complexity of doing the work is really that there are

38:30 so many elements at hands both on the regulatory point of side,

38:34 on the firm specific point of file side,

38:37 the counter-specific

38:40 point that uh that actually you,

38:42 you really need to,

38:44 to

38:44 go very granular in terms of

38:47 um

38:48 of uh uh of data and uh.

38:51 And information and that's why I was thinking that

38:55 working with a survey may be a useful compliment.

39:00 And maybe if I still have

39:02 one second after this long introduction to explain why I'm showing this

39:09 to you,

39:09 I will try to share a presentation where I will just

39:14 show a couple of slides,

39:18 um.

39:20 And I think,

39:20 uh,

39:20 can you see it?

39:22 Yes,

39:22 it looks good.

39:24 OK,

39:25 and I think I have to,

39:28 OK.

39:29 Um,

39:30 what I will show you something that,

39:32 uh,

39:32 we launched,

39:33 uh,

39:34 7 years ago at the European Investment Bank is a

39:36 survey of more than 12,000 firms at the European level,

39:41 representative for single countries and with a benchmark at the European,

39:46 uh,

39:46 at the US level

39:48 and also is the links to balance sheet data of the firms.

39:52 So we combine the balance sheet information of the firms.

39:56 With uh

39:57 qualitative and quantitative data that uh we

40:01 um

40:02 we uh gain from uh from uh the survey

40:06 exercise

40:07 and what we can do is,

40:09 uh,

40:09 uh,

40:10 first of all ask the firms to think about the barrier

40:13 on investment here in particular in the NASA survey we had a

40:18 very strong rule of energy costs as a

40:20 barrier to investment and uncertainty at the European.

40:24 Level

40:24 and that's allowed us also combined to the other

40:29 variables to understand what was happening in terms of

40:33 climate investment and energy efficiency investment in Europe.

40:37 So what we find out is exactly that

40:41 in the moment in which in Europe we had the

40:44 energy crisis last year coming after the war in Ukraine,

40:48 what we had was that

40:51 Energy costs were a driver for

40:55 investment in energy efficiency,

40:57 but on the other side,

41:00 the

41:02 uncertainty was

41:03 was a drag to investment overall.

41:06 But when we look at energy efficiency,

41:08 the net effect

41:10 of

41:11 the push for the energy cost and the uncertainty

41:15 left a small positive.

41:17 So increasing the probability.

41:19 Ability for firms

41:20 to invest in energy efficiency and in fact we see it at the European level with

41:27 very strong investment in energy efficiency this

41:29 year in the private sector coming up.

41:32 What happened with the climate action investment that is a wider concept

41:37 not only energy efficiency but also adaptation and all other element of mitigation

41:43 was in fact that the

41:45 um

41:47 Positive impact of the energy cost was a

41:50 completely balanced by the fact of uncertainty.

41:54 And we explain it also because on a policy

41:56 point of view there were various instruments that were

42:00 um

42:01 Used also to contain energy prices to

42:05 um

42:06 to,

42:06 to support the firms during the crisis

42:10 that uh distorted incentive on investment.

42:13 So

42:14 what you had is that the combination of

42:16 uncertainty and energy costs left a positive stimulus

42:20 for energy efficiency but it didn't change the

42:23 probability of firms to invest in climate action.

42:26 And that's something important uh to understand also on a policy point of view,

42:31 to know which kind of instrument and policies that work

42:35 to stimulate the climate investment action.

42:39 And then uh uh an example of what we are doing in the survey,

42:43 we are asking

42:44 firms what they do for mitigation and for adaptation

42:49 and also their perception,

42:50 of course,

42:51 uh,

42:51 before I don't show it of physical risk and transition risk.

42:55 What you see on mitigation,

42:57 we ask them which kind of measure they implement.

43:01 You see a lot of use of energy efficiency

43:04 investment of waste minimization that helps us understanding.

43:08 The regulation of matters

43:10 and actually in Europe we see the effect of a regulation

43:13 in imposing a waste minimization for firms and that's why that's

43:17 an area so developed.

43:20 But it's also interesting that the um

43:23 the investment in new less polluting technologies or activities which is

43:28 the transformation of firms moving

43:31 either changing technologies or moving in

43:34 business areas and business activities.

43:37 That are less polluting.

43:39 What we see there and it's the current work that we are developing,

43:43 we are trying to look at the most energy intensive sectors

43:47 and looking at what motivates the firms that are actually

43:52 investing to transform and those that are

43:55 not transforming

43:56 and with more binding regulation coming in,

44:00 you would expect that those that are non-transforming,

44:04 they are trying to use the assets.

44:06 In the end,

44:06 and then they will just close the activities

44:09 while uh the others uh start transforming the business already now.

44:14 I think uh

44:15 further analysis and further work in this direction is very

44:19 important because of the structural shift that the economy will have

44:24 and understanding how to,

44:26 um,

44:28 how the reallocation of resources and also

44:31 the transformation of firms that will happen is extremely important.

44:35 With that,

44:36 uh,

44:36 I think,

44:36 uh,

44:37 uh,

44:37 I had in the presentation a couple of additional slides to show a project that

44:42 we developed actually also with uh Jorge

44:45 and with DBRD based on the enterprise survey

44:49 both on the North African region and on the East and North Africa second project.

44:57 There as well we were looking at the physical climate risk,

45:01 green investment and trying to understand that the

45:04 uh drivers for firm transformation.

45:07 And again I think that was an early

45:10 um attempt with an early version of a green model for the enterprise survey.

45:15 I think that going uh um.

45:19 The more direct

45:21 as possible,

45:22 as direct as possible with the new set of definition that now

45:27 start becoming uh um standard and understood

45:30 also by the corporate sector in terms of

45:33 what is the physical risk,

45:35 what is the transitional risk,

45:36 and what can be done in a mitigation and adaptation.

45:39 Having a more spelled out question I would be.

45:43 Useful and would help also uh around the world

45:46 a better understanding of what motivates the film.

45:49 I hope this uh um

45:51 helps in the discussion and uh

45:55 and uh that I didn't bring uh topics completely out

45:59 of uh of uh what you wanted to discuss,

46:02 but I think that the climate part is very important to consider.

46:10 Thanks very much,

46:11 Deborah.

46:12 Um,

46:12 uh,

46:13 Chris,

46:13 here,

46:14 uh,

46:14 our last panelist.

46:15 Go ahead.

46:16 Great.

46:16 Thank you.

46:17 And,

46:17 and thanks for,

46:18 uh,

46:18 the chance to,

46:19 to,

46:19 to,

46:20 to speak to you today.

46:21 Um,

46:21 first,

46:22 I wanna say congratulations,

46:23 uh,

46:23 to,

46:24 to KCP for,

46:25 for the work you've done over the years and,

46:27 and the amount of time.

46:28 I think

46:29 I wanna sort of both reflect back a bit over that period of time and what we've,

46:33 I think,

46:34 accomplished and also kind of think about

46:36 what

46:36 really is

46:38 is maybe where we've done less,

46:40 where we've been,

46:40 where we've made less progress,

46:42 and where I think the next 20 years,

46:44 uh,

46:44 might,

46:44 uh,

46:45 uh,

46:45 might,

46:45 might focus.

46:46 And,

46:46 and as I sort of reflected back at,

46:49 at,

46:49 at,

46:50 on the 20 years,

46:50 I,

46:51 I thought about the work that David and I started doing,

46:54 you know,

46:54 I guess a bit more than 20 years ago.

46:57 Um,

46:57 and thinking about

47:00 Um,

47:00 data

47:01 and microenterprises and where we were with kind of how we understood

47:06 how,

47:06 what profits were,

47:07 how firms,

47:09 um,

47:09 thought about

47:10 investments,

47:11 thought about returns and things,

47:13 uh,

47:13 things of that sort in places where

47:15 people,

47:16 uh,

47:16 didn't even,

47:17 uh,

47:17 keep,

47:17 uh,

47:18 any,

47:18 any written records.

47:20 Um,

47:20 and the literature was really quite

47:22 sparse in a sense.

47:24 Um.

47:24 I think

47:25 one of the,

47:26 if,

47:26 if I,

47:27 I think about that time,

47:28 that one of the surprising

47:29 um pieces of data that was out there was the,

47:33 the,

47:33 uh,

47:34 the Enemin survey in,

47:35 in Mexico,

47:36 the microenterprise survey in Mexico,

47:37 which,

47:37 which David and I used early,

47:39 early on,

47:40 um,

47:41 which was in some senses,

47:42 way,

47:42 way ahead of its time.

47:43 And if I think about the 20 years and where we've come from,

47:47 from that point in time,

47:48 um,

47:49 From the perspective of data

47:51 and how we understand these firms,

47:53 it's really quite amazing,

47:55 uh,

47:56 amazing progress.

47:57 And,

47:58 uh,

47:58 And,

47:59 uh,

47:59 as,

47:59 as was pointed out,

48:00 a lot of this work,

48:01 you know,

48:02 uh,

48:02 important work was funded by,

48:03 uh,

48:04 you know,

48:04 by the bank and by,

48:06 by KCP,

48:07 um,

48:07 but it,

48:08 but we're,

48:09 we're at a point where it's,

48:10 we're way beyond kind of,

48:12 uh,

48:13 measuring,

48:14 um,

48:15 uh,

48:16 what,

48:16 what firm,

48:16 how firms keep,

48:18 how,

48:18 how firms think about profits and,

48:20 and,

48:20 and,

48:21 and,

48:21 uh,

48:21 and other things,

48:22 and thinking about

48:23 really using technology and using experiments and using,

48:27 um,

48:28 Uh,

48:28 other kinds of,

48:29 uh,

48:30 information to think about,

48:31 um.

48:32 Uh,

48:33 effort at,

48:34 at,

48:34 at,

48:35 at their,

48:35 when,

48:35 when they're working in the enterprise and other aspects of,

48:39 of,

48:39 of productivity,

48:41 um,

48:42 that is,

48:43 uh,

48:43 is,

48:44 is,

48:44 you know,

48:44 really quite some,

48:46 some startling progress.

48:47 And I,

48:47 and I think the progress has been

48:49 most,

48:50 uh,

48:51 uh,

48:51 impressive and,

48:52 and strongest in the area of,

48:54 of microenterprises and thinking about

48:57 these very small scale,

48:58 uh,

48:58 these very,

48:59 very small scale,

49:00 uh,

49:00 enterprises.

49:01 I think recently,

49:02 there's also been a fair amount of work in very large enterprises.

49:06 And,

49:07 um,

49:08 and in,

49:08 in a way,

49:09 these are

49:10 the,

49:10 you know,

49:11 the,

49:11 what's,

49:12 what's missing is,

49:13 is the missing middle.

49:14 And it's not just cause the firms,

49:16 there are not as many firms there,

49:17 but because we,

49:18 we haven't,

49:19 I think,

49:20 focused as,

49:21 with some notable exceptions,

49:22 haven't focused as much attention on,

49:25 on sort of mid-sized firms.

49:26 And

49:27 It's kind of obvious as researchers why that's the case,

49:30 that we end up,

49:31 uh,

49:31 going to the field and seeing

49:33 massive numbers of very small-scale firms.

49:36 It's easy to get a sample,

49:37 we can do,

49:38 we can do work,

49:39 or we can go develop a relationship with a large-scale firm,

49:42 and the firm's large enough that we can do

49:45 experiments and other work,

49:46 uh,

49:47 within a single or a handful of,

49:49 of,

49:49 of very large firms.

49:51 Um,

49:51 but if I think about the kinds of firms that are,

49:55 um,

49:55 participating in

49:57 incubators,

49:57 accelerators,

49:58 the sort of fast-growth,

49:59 you know,

50:00 sort of,

50:00 uh,

50:01 uh,

50:02 uh,

50:03 mid-sized firms,

50:05 um,

50:05 there's,

50:06 I think,

50:07 quite a lot,

50:08 you know,

50:08 less work in that area.

50:10 And I think that's a place where,

50:12 um,

50:13 where

50:13 we need,

50:14 uh,

50:14 we,

50:15 you know,

50:15 we need,

50:15 uh,

50:16 we need more attention.

50:18 And I wanna,

50:19 I,

50:19 I guess I wanna say,

50:20 I think,

50:21 um,

50:21 and,

50:22 and

50:23 Jorge,

50:23 you,

50:23 you mentioned the difficulty with,

50:25 with administrative data and the lack of administrative data,

50:28 and I think

50:29 that's right.

50:30 There's often in many countries,

50:32 a lack of administrative data,

50:33 but we're beginning to see

50:35 now

50:36 in,

50:36 in some quite,

50:37 you know,

50:38 in some quite important

50:39 instances of people getting access to

50:43 value-added transaction tax data,

50:46 to trade data,

50:47 to uh corporate income tax data,

50:49 uh,

50:49 and being able to sort of merge across data sets

50:53 uh in India,

50:54 and Pakistan,

50:55 and Uganda,

50:55 and Kenya,

50:56 and other,

50:57 in other countries.

50:58 And

50:59 those

51:00 Data,

51:01 I think,

51:01 have

51:02 uh,

51:02 a tremendous amount of promise

51:05 in

51:06 helping us understand

51:07 um

51:08 both firm dynamics generally

51:10 and the sorts of general equilibrium effects that,

51:12 that,

51:13 uh,

51:13 that,

51:13 that Dean,

51:14 that Dean was,

51:14 was talking about.

51:16 Um,

51:16 they have the advantage of being comprehensive.

51:18 They have the advantage of being,

51:20 um,

51:21 Uh,

51:21 you know,

51:22 available over a long period of time.

51:24 Um,

51:25 of course,

51:25 only for formal,

51:26 for formal firms.

51:28 But,

51:28 uh,

51:29 but the,

51:30 these,

51:30 you know,

51:31 these sort of faster growing mid-sized firms

51:33 are,

51:33 are likely,

51:34 uh,

51:34 to be.

51:34 And so,

51:35 I think the

51:37 The

51:38 one of the challenges we face as,

51:40 as researchers and,

51:42 and,

51:42 and,

51:43 and when,

51:43 when working with,

51:44 with policymakers is

51:46 figuring out ways

51:47 to,

51:48 to make those,

51:49 uh,

51:49 those kinds of data more accessible.

51:52 And,

51:52 uh,

51:53 for example,

51:53 Uganda is setting up a secure data center.

51:56 South Africa has,

51:57 has done some work in setting up secure data

51:59 centers to make the data more broadly available.

52:02 I mean,

52:03 I think over

52:04 The past,

52:05 they've been

52:06 typically available when you get the right relationship and you some,

52:09 and you,

52:10 and you get access to the,

52:11 to the data.

52:12 But,

52:12 uh,

52:13 but I think

52:14 the challenge of making those more systematically

52:17 available is one that I think the bank can,

52:19 uh,

52:20 can,

52:20 can play a,

52:21 can play a role in,

52:21 and I think would have,

52:22 uh,

52:23 uh,

52:24 if you build it,

52:24 they will,

52:25 they will come kind of sense of generating a lot of research,

52:28 uh,

52:28 uh,

52:29 a lot of research in this,

52:30 um,

52:31 in,

52:31 in this area.

52:32 And

52:34 Uh,

52:34 one of the things that I would say,

52:36 I again see as an area that's developing now,

52:40 and

52:41 an area that's likely to grow is exactly on the sort of,

52:44 uh,

52:45 um,

52:46 interaction between,

52:48 uh,

52:49 microeconomists and macroeconomists,

52:51 um,

52:52 or at least

52:52 the interest in macroeconomists of,

52:55 uh,

52:55 in,

52:55 in the private sector and in firms.

52:57 Um.

52:58 I run a,

52:59 uh,

53:00 some of you will know,

53:01 I run a program

53:02 funded by uh FCDO,

53:05 uh,

53:05 uh,

53:06 called Private Enterprise Development in Low-income Countries,

53:08 a grants program.

53:09 There's a sister program called Structural Transition,

53:12 uh,

53:12 uh,

53:13 and Economic Growth,

53:14 uh,

53:15 which is basically a macro program.

53:17 And SEG funds

53:18 really quite a substantial amount of

53:21 work on firms and work on the private sector,

53:23 and much of it is exactly in this area of kind of thinking about,

53:27 um,

53:27 Uh,

53:28 you know,

53:29 how we think about general equilibrium effects,

53:31 how we think about using,

53:33 uh,

53:33 structural models,

53:35 um,

53:35 uh,

53:36 to,

53:36 uh,

53:37 to,

53:37 to,

53:38 to,

53:38 to come up with sort of policy counterfactuals.

53:41 We're not quite there yet in,

53:42 in marrying the,

53:43 the microeconomics with the,

53:45 with the macroeconomics,

53:46 but,

53:47 but there have been

53:48 Really important moves in the,

53:49 in that direction in the,

53:50 in the last,

53:51 just,

53:52 you know,

53:52 recent

53:53 years.

53:53 And I think that's also a place where,

53:56 um,

53:56 where there's a lot of potential,

53:58 uh,

53:58 in the,

53:59 in,

53:59 in the next,

54:00 uh,

54:01 in the next,

54:01 in the next 20 years.

54:02 So,

54:03 I'll,

54:03 I'll stop,

54:04 I'll leave it at that and,

54:05 and say,

54:06 I think,

54:06 you know,

54:08 Measurement data,

54:09 uh,

54:10 and David mentioned the management practices.

54:12 Another area where,

54:14 um,

54:15 uh,

54:15 you know,

54:16 uh,

54:16 as,

54:16 as the saying in management goes,

54:18 what gets measured gets managed.

54:20 You know,

54:20 when,

54:21 when we're able to measure things within,

54:23 within,

54:23 uh,

54:24 Uh,

54:25 within these firms and across these firms,

54:27 um,

54:28 you know,

54:28 we were able to,

54:30 to make much more progress in,

54:32 in thinking about what policies are and what,

54:34 what interventions are that,

54:35 that will,

54:36 uh,

54:36 that will,

54:37 will benefit

54:38 firms and,

54:38 and,

54:39 and firm growth.

54:39 So,

54:41 Data measurement,

54:42 yes,

54:43 the way we've been doing it with surveys and with,

54:45 and with developing kind of new,

54:47 um,

54:48 relatively small sample,

54:50 uh,

54:50 rich data sets.

54:52 But I think also data,

54:54 uh,

54:54 thinking more about,

54:55 uh,

54:56 the administrative data and the,

54:57 and the,

54:58 the ability to kind of

55:00 Reach much larger populations of firms,

55:02 uh,

55:03 where

55:04 firms that are less common when we go to the field will,

55:07 will,

55:07 will be,

55:08 uh,

55:08 available and,

55:09 and,

55:10 uh,

55:10 and there in,

55:11 in,

55:11 in larger numbers,

55:12 uh,

55:13 uh,

55:13 will be,

55:14 uh,

55:14 will be important as we go,

55:15 as we go forward.

55:16 And I'll,

55:17 I'll leave it there.

55:23 Thanks very much,

55:24 Chris,

55:24 and uh I'm not sure if you're able to connect to the first part of the session,

55:27 but you're really kind of bringing us full circle because

55:29 we had a lot of discussion about the importance of,

55:32 of drawing on administrative data,

55:34 um,

55:35 you know,

55:35 both on firms,

55:36 on tax records,

55:37 and so on as a way of,

55:38 uh,

55:39 of generating insights and also the,

55:41 the difficulties in accessing this data.

55:43 So we have,

55:44 excuse me,

55:45 we have about 20 minutes um remaining uh for questions.

55:49 Um,

55:50 so the floor is open,

55:51 we'll start with some in the room,

55:52 and if we have any online,

55:53 please,

55:53 please let me know,

55:54 Karina.

55:55 So,

55:55 uh,

55:55 Sergio and then Daria,

55:57 and then Karina if we have some people online.

56:00 For us.

56:01 And,

56:01 uh,

56:01 thanks for the presentation.

56:02 It's very interesting.

56:03 Um,

56:04 one thing for advertising and,

56:06 um,

56:06 following what Chris mentioned is at the macro group and also using KCP funds.

56:10 We are trying to

56:11 link the micro and the macro

56:13 aspects of the economy using,

56:15 um,

56:16 more widely available data at the country level or cross country.

56:20 But,

56:20 uh,

56:21 my question was,

56:21 uh,

56:22 basically to David and,

56:24 and Dean,

56:25 you mentioned,

56:26 uh,

56:26 this issue of a black box of what entrepreneurship,

56:29 uh,

56:30 abilities are.

56:31 Um,

56:31 and you mentioned the value of,

56:33 uh,

56:33 doing consulting and training,

56:35 uh,

56:35 to these entrepreneurs that can increase performance over time.

56:40 But it's still,

56:41 for the audience,

56:42 that's a black box.

56:44 So,

56:44 I was wondering whether you can uh provide a bit more examples

56:48 of what are the practices that you use to train these entrepreneurs

56:53 and what are the most effective uh training practices that you have uh seen

56:57 that have an effect on,

56:58 on,

56:59 on entrepreneurial activity later on.

57:05 Um,

57:06 yes,

57:06 my question,

57:07 um,

57:07 I think really to whoever wants to pick it up,

57:10 but,

57:11 uh,

57:11 so there is,

57:12 there's been a lot of discussion about what happens within the firm,

57:15 but from the trade side,

57:17 we see that really this firm to firm linkages are very important.

57:20 So I was wondering whether there is,

57:22 uh,

57:23 um,

57:23 you know,

57:24 a program there to perhaps bridge that dimension.

57:30 Now,

57:31 I wanted to,

57:31 I mean,

57:31 it's maybe more common than the question,

57:33 but um I think,

57:34 you know,

57:35 the,

57:36 the link between blue sky type of,

57:38 of research and then policy impact as,

57:40 as uh David really,

57:41 uh,

57:42 you know,

57:42 outlined very well and uh documented well,

57:45 is really,

57:45 you know,

57:45 I think when you look at David's presentation,

57:47 he really schemes

57:49 uh through all the important steps in between getting that first,

57:53 you know,

57:54 uh,

57:54 paper that was academically received.

57:56 But then making those linkages with policymakers and really,

58:00 you know,

58:00 it's not like you have a paper in the

58:02 top five journal and suddenly people hear your messages

58:05 in the circles of influence,

58:06 right?

58:06 So,

58:07 I think kind of

58:08 thinking about together about the,

58:09 the importance of KCP financing and the work of

58:11 the bank more generally in making these linkages and,

58:14 and can we

58:16 think a little bit differently about KCP financing and making

58:18 this work a little bit easier for researchers to really

58:21 uh make that link.

58:25 So,

58:25 there's a question from Alvaro Gonzalez,

58:28 which is actually similar to what Sergio just asked,

58:30 and how can we focus more on

58:33 the firm,

58:33 the kind of firm dynamics that can promote better uh managed firms,

58:38 and

58:38 how do we better identify factors that really improve the quality of management.

58:42 This is more for David.

58:44 Thanks.

58:47 Anybody else in the room wanna come in?

58:49 Norman,

58:49 please.

58:54 Thank you.

58:55 Uh,

58:55 great presentation,

58:57 David,

58:57 Jorge.

58:58 Uh,

58:59 I want to

59:00 make a comment on,

59:01 on administrative data,

59:03 uh,

59:04 that has been the subject of the,

59:05 of this session and the previous one.

59:08 And,

59:08 uh,

59:09 present maybe a counterpoint

59:11 on,

59:11 uh,

59:12 on the use of administrative data.

59:14 I,

59:15 I do wonder in some,

59:17 for some countries,

59:18 whether this data

59:20 are truly reliable.

59:23 Um,

59:23 and whether we should be concerned about

59:26 that reliability.

59:27 Uh,

59:28 and if that is a concern,

59:30 then how can we

59:32 Combine

59:34 administrative data with

59:36 independently collected data

59:38 like we do in the,

59:39 the Enterprise survey program

59:41 to uh improve the

59:45 reliability of such data.

59:47 And the second point that is related

59:50 is,

59:50 uh,

59:51 well,

59:51 some countries are making this data

59:54 available,

59:55 but then,

59:56 can we extrapolate

59:57 to other countries

59:58 that don't make the data available

1:00:01 for some reason.

1:00:03 Um,

1:00:04 and

1:00:05 my conjecture is that

1:00:07 extrapolation is very hard.

1:00:09 And that's why we also need independent data collection.

1:00:12 So,

1:00:13 2 issues of administrative data.

1:00:15 Thank you.

1:00:18 Oh,

1:00:18 thank you very much,

1:00:18 Norman.

1:00:19 I was actually going to raise uh perhaps even a sharper version of that question,

1:00:22 because it's not just a question of reliability,

1:00:24 it's also a question of,

1:00:26 in some cases,

1:00:27 political influence,

1:00:28 agendas,

1:00:28 and so on,

1:00:29 uh,

1:00:30 that we have to find ways to filter through with administrative data,

1:00:33 and I,

1:00:34 I think,

1:00:34 well,

1:00:34 particularly in your be Ready project,

1:00:37 you're really trying to navigate that,

1:00:38 um.

1:00:39 Uh,

1:00:40 let,

1:00:40 let me add,

1:00:41 uh,

1:00:42 two other quick questions to the mix,

1:00:44 and then we'll go back to the,

1:00:45 uh,

1:00:45 the speakers and the panelists for,

1:00:47 you know,

1:00:48 1 or 2 minutes each for,

1:00:49 uh,

1:00:49 for responses.

1:00:51 Um,

1:00:52 uh,

1:00:52 I mean,

1:00:52 I'm,

1:00:52 I'm picking up a little bit on

1:00:54 sort of the,

1:00:54 the similar question that,

1:00:56 that Dean posed.

1:00:57 Um,

1:00:58 uh,

1:00:58 so Dean asked the question about,

1:01:00 you know,

1:01:00 why is it that consulting firms aren't getting rich?

1:01:02 And I think the flip side of that question is,

1:01:05 You know,

1:01:06 why isn't management sort of like

1:01:08 just another input like capital and labor,

1:01:10 and why are the sort of puzzles different in,

1:01:13 in that case?

1:01:14 Like,

1:01:14 you know,

1:01:15 for example,

1:01:15 one could take the view that firms don't invest in management because

1:01:19 they're small,

1:01:19 and one might even make a case that there's some kind of a,

1:01:22 you know,

1:01:22 size trap mechanism that's driving that.

1:01:25 Although I know from a paper that you and I wrote together,

1:01:27 David,

1:01:27 that we're not very sympathetic to that view.

1:01:29 So,

1:01:29 if it's not that trap view,

1:01:31 uh,

1:01:31 what would the other view be?

1:01:33 Um,

1:01:33 and I have a question for,

1:01:34 for Deborah as well.

1:01:35 So you showed some very interesting stuff on the green side,

1:01:38 and I think you made a very compelling case that

1:01:40 thinking about green incentives for firms are important.

1:01:43 One question I had seeing the slides from your EIBIS

1:01:47 survey is,

1:01:49 you know,

1:01:49 what do we learn from these questions about

1:01:52 sort of

1:01:53 private incentives versus external benefits of

1:01:57 investments that firms are making,

1:01:59 right?

1:01:59 Because the trite answer always when

1:02:01 we sort of confront that question and say,

1:02:03 Well,

1:02:03 your governments should just put carbon taxes and,

1:02:05 right?

1:02:06 But,

1:02:06 but we know that governments can't put carbon taxes.

1:02:08 So,

1:02:08 we really need to understand,

1:02:10 well,

1:02:10 in the absence of carbon taxes,

1:02:12 what are,

1:02:13 what are the things that are actually in firms' private interest to do

1:02:16 and what are the things that

1:02:17 they're doing insufficiently because of externalities,

1:02:20 and how can these surveys tell us more about that?

1:02:23 So,

1:02:24 with that,

1:02:24 let's go first to Jorge and David,

1:02:27 um,

1:02:28 and then we'll go to the panelists,

1:02:29 and if I could ask you to be a bit disciplined about sticking to,

1:02:32 you know,

1:02:32 2 minutes each,

1:02:33 we'll be able to wrap up on time.

1:02:36 Yes,

1:02:36 a few of the topics perhaps,

1:02:37 uh,

1:02:38 addressing first the,

1:02:38 the issue of the um initiative data

1:02:42 uh for uh Christopher,

1:02:43 I think he

1:02:44 raised that and Norman's point.

1:02:46 Uh,

1:02:46 the,

1:02:47 the,

1:02:47 the,

1:02:47 I think I agree with Christopher in the sense that,

1:02:49 uh,

1:02:49 yes,

1:02:49 you do,

1:02:50 we do have a big gain access to some very good data in some specific countries.

1:02:55 The point I was making,

1:02:56 uh,

1:02:56 is actually to the point he said that word systematic.

1:02:58 Around the globe.

1:02:59 It's,

1:03:00 I don't think this is a very promising.

1:03:01 It's,

1:03:01 it's kind of like Nietzsche,

1:03:02 what we can find in some certain countries,

1:03:04 great,

1:03:05 but when we try to think of that around the world,

1:03:07 it's very limited.

1:03:09 Uh,

1:03:09 and then there's restrictions.

1:03:10 I am interested in that among the countries that you mentioned,

1:03:13 one of those is one of the countries in which I have my faced most difficulties

1:03:17 accessing any data for,

1:03:18 for any kind of work,

1:03:19 for research

1:03:20 and,

1:03:20 and,

1:03:21 and,

1:03:21 and.

1:03:21 And even for the implementary services in the case,

1:03:23 the case of South Africa.

1:03:24 Um,

1:03:25 and I do know that we gain access to that access just to use as an example,

1:03:28 we have gained access to that,

1:03:29 but a very restrictive way of using the data and the bank has,

1:03:33 has profited from that.

1:03:34 But the point I was making is that it's a little bit

1:03:36 uh difficult to put your money into that,

1:03:38 into,

1:03:39 into,

1:03:39 uh,

1:03:39 on a global basis and I realized we,

1:03:42 the World Bank probably should play a role on that.

1:03:44 Uh,

1:03:44 uh,

1:03:45 and,

1:03:45 uh,

1:03:45 it,

1:03:45 it,

1:03:46 it has come to me several times,

1:03:47 the request of me playing the role in that way

1:03:49 is because I interact with a lot of these,

1:03:50 uh,

1:03:50 with all the countries around the world in

1:03:52 my role as manager of interpersonallysis unit,

1:03:54 and at least a little bit of my frustration that I was expressing is,

1:03:57 uh,

1:03:57 my experience from,

1:03:58 from building from all those interactions in which is,

1:04:00 uh,

1:04:00 is,

1:04:01 is,

1:04:01 is,

1:04:01 it's,

1:04:01 it's not,

1:04:02 it,

1:04:02 it,

1:04:02 you just,

1:04:03 we just,

1:04:03 um,

1:04:04 bump into issues of legality and it's not nothing researchy,

1:04:07 it's poli,

1:04:07 it's purely political,

1:04:08 the access to this administrative data.

1:04:10 Um.

1:04:12 I think,

1:04:12 uh,

1:04:12 on,

1:04:12 on,

1:04:13 on,

1:04:13 on,

1:04:13 I just wanted to touch base on,

1:04:14 on,

1:04:14 on Deborah's point as well on the green practices in the green of firms

1:04:18 in,

1:04:19 uh,

1:04:19 in the,

1:04:20 the,

1:04:20 the,

1:04:20 she made the case of

1:04:22 the work we did together in the past

1:04:23 and,

1:04:23 uh,

1:04:24 uh,

1:04:24 but she was making the case of uh trying to build more into the service,

1:04:28 uh,

1:04:28 this transition of the motivation of the firms into the transition.

1:04:31 I fully agree and I would be happy to explore

1:04:33 how they're doing it and see how we explore,

1:04:35 but I also So,

1:04:36 I would add,

1:04:36 moving to the table that I would like to discuss how,

1:04:38 what is the best way to,

1:04:39 way to do that?

1:04:40 Is it survey data,

1:04:41 the best way to address that or is there other

1:04:43 methods of data collection that probably would be better.

1:04:45 I'm not very keen on asking direct questions about incentives

1:04:49 to firms because I'm more keen on observing how they,

1:04:51 how they act

1:04:52 in trying to do that because

1:04:54 um this,

1:04:55 those direct questions on,

1:04:57 on or perceptions in a sense,

1:04:58 uh can be misleading.

1:05:00 Um,

1:05:00 I think just a little bit on,

1:05:02 on,

1:05:02 I think,

1:05:02 uh,

1:05:02 on the,

1:05:03 on the,

1:05:03 within the firm.

1:05:04 There's a little bit of work on,

1:05:05 on my,

1:05:05 on my team,

1:05:06 more on the research side,

1:05:06 not on data collection within the firm on that,

1:05:08 on what's happening within the firm,

1:05:10 uh,

1:05:10 but I,

1:05:11 but it's true.

1:05:11 It's not,

1:05:11 not much systematic in the,

1:05:13 at least on the data collection effort that we do have.

1:05:15 It's a little more on the,

1:05:16 on the,

1:05:17 on the research side of,

1:05:18 of a couple of people in my team.

1:05:22 So,

1:05:23 um,

1:05:23 yeah,

1:05:24 thanks,

1:05:24 thanks to all these great,

1:05:25 um,

1:05:26 comments and,

1:05:27 and discussing comments.

1:05:28 Uh,

1:05:29 so let me just pick up on a couple of things,

1:05:30 and so,

1:05:31 one is this,

1:05:32 um,

1:05:33 point we sort of get all the time,

1:05:34 which is,

1:05:36 We,

1:05:36 you know,

1:05:36 we work for years,

1:05:38 we,

1:05:38 we

1:05:39 do a big sample,

1:05:40 we,

1:05:40 we painstakingly look at this data,

1:05:42 and if we,

1:05:43 and,

1:05:43 you know,

1:05:44 a lot of the times we find things that don't work,

1:05:45 and then when we find something that works,

1:05:47 The question is always,

1:05:48 well,

1:05:48 if it works,

1:05:49 why isn't the private sector already doing this already,

1:05:51 and,

1:05:51 you know,

1:05:52 I had to,

1:05:53 you know,

1:05:53 interview

1:05:54 3000 firms,

1:05:55 trace them over,

1:05:57 you know,

1:05:57 3 years,

1:05:58 um,

1:05:58 do a lot of econometrics,

1:05:59 and I can barely detect the impact,

1:06:01 and then we think that everybody should know that this is,

1:06:04 you know,

1:06:04 what,

1:06:05 what's going on.

1:06:05 And so,

1:06:06 um,

1:06:07 you know,

1:06:07 I think

1:06:08 it,

1:06:09 you know,

1:06:09 a lot of this stuff,

1:06:10 it's,

1:06:10 it's actually,

1:06:11 even when you've gone through this,

1:06:13 there's just so much other stuff that hits firms that it's actually very hard.

1:06:16 Hard for them to know whether,

1:06:18 um,

1:06:18 what they're doing works or,

1:06:20 or not.

1:06:20 And so,

1:06:20 this is,

1:06:21 I think,

1:06:21 one of the big failures in the consulting market is,

1:06:23 is that

1:06:24 it's,

1:06:25 um,

1:06:25 you know,

1:06:25 it's an experience,

1:06:26 good.

1:06:26 You don't know what you're getting before you go in,

1:06:28 as,

1:06:29 as Sergio said,

1:06:30 you know,

1:06:30 you told me about consulting,

1:06:31 but I don't really know what that,

1:06:33 that is.

1:06:33 Um,

1:06:34 you know,

1:06:34 there's a sort of set of practices,

1:06:36 like,

1:06:36 you know,

1:06:37 even at a general level,

1:06:38 if I say,

1:06:38 well,

1:06:38 it's about quality improvement and logistics improvement and HR practices,

1:06:41 well,

1:06:41 that's great,

1:06:42 but then what specific practices,

1:06:44 and,

1:06:44 you know,

1:06:44 that's all very context-specific.

1:06:46 Um,

1:06:46 so,

1:06:47 I think it's,

1:06:47 you know,

1:06:48 very hard for firms to know what they're getting ex ante.

1:06:50 We do see that when firms have gone through this,

1:06:52 they're more likely to go back and buy a bit of

1:06:55 this consulting on the market themselves.

1:06:56 So I think there is this sort of experience,

1:06:58 good

1:06:59 aspect.

1:06:59 But I think it's just really hard,

1:07:01 um,

1:07:01 you know,

1:07:01 there's this great paper I like on advertising.

1:07:04 Where,

1:07:05 um,

1:07:05 firms that are doing experiments with 2 million customers

1:07:08 can't tell whether they're getting 50% return or 0% return

1:07:12 on that advertising expenditure.

1:07:13 It's just like super hard to know when so much other stuff hits your firm,

1:07:17 whether anything you do,

1:07:19 do works or not.

1:07:20 And so,

1:07:20 I think this is like,

1:07:21 you know,

1:07:22 one of the big challenges for us.

1:07:23 And so that's where,

1:07:23 you know,

1:07:24 the,

1:07:24 the sort of macro factors of,

1:07:26 um,

1:07:27 You know,

1:07:27 what's driving firm dynamics and,

1:07:30 and,

1:07:30 you know,

1:07:30 what else can we have at that sort of ecosystem level in terms of competition policy,

1:07:34 and some of our

1:07:35 colleagues have been working on that,

1:07:37 and,

1:07:37 and,

1:07:38 uh,

1:07:38 um,

1:07:39 you know,

1:07:39 what can

1:07:40 large firms do through their value chains of trying to help

1:07:44 their,

1:07:44 their customers improve management and,

1:07:46 you know,

1:07:46 we see trade and having this role in improving.

1:07:48 Um,

1:07:49 and upgrading as well.

1:07:50 And so I think it's,

1:07:50 you know,

1:07:51 trying to look from the bottom up and,

1:07:52 and the top down,

1:07:53 but I think the sort of point of,

1:07:54 you know,

1:07:54 when you finally find something that maybe works,

1:07:57 um,

1:07:57 why has everyone not done it already?

1:07:58 Uh,

1:07:59 and yet,

1:07:59 you know,

1:08:00 I,

1:08:00 I do a lot,

1:08:01 we also do lots of well-intended things that don't work,

1:08:03 um,

1:08:04 is,

1:08:04 is there.

1:08:04 So let me stop there and I'll,

1:08:06 um,

1:08:06 you know,

1:08:06 I'm sure there's other questions,

1:08:07 but I wanna give the panelists time to respond.

1:08:10 OK,

1:08:11 thanks.

1:08:11 Let's go to the panelist's kind of in reverse order.

1:08:13 So,

1:08:14 um,

1:08:15 uh,

1:08:15 Chris,

1:08:15 do you have a couple of observations?

1:08:18 Yeah,

1:08:19 sure.

1:08:19 So,

1:08:19 so a couple of things.

1:08:20 Uh,

1:08:21 one is,

1:08:21 one is on the question of,

1:08:23 um,

1:08:23 on what works in training and,

1:08:25 and,

1:08:25 and so forth.

1:08:26 I,

1:08:26 I wanted to bring it back to,

1:08:27 to one of the slides that,

1:08:28 that Jorge had about,

1:08:30 um,

1:08:30 uh,

1:08:31 about gender and discrimination and so forth and say that

1:08:33 I think that,

1:08:34 um,

1:08:36 There's one issue is what you can do to sort

1:08:38 of improve entrepreneurial ability among

1:08:40 people who are currently entrepreneurs,

1:08:41 but I also think we shouldn't lose sight of the fact that

1:08:44 we,

1:08:44 we,

1:08:44 and a lot of these economies,

1:08:45 we're not doing a very good job of

1:08:47 selecting who becomes an entrepreneur to begin with.

1:08:49 And we need to think hard about that,

1:08:51 about that,

1:08:52 uh,

1:08:52 that,

1:08:52 uh,

1:08:53 question as,

1:08:53 as,

1:08:54 as,

1:08:54 as well.

1:08:55 And,

1:08:55 and,

1:08:55 um,

1:08:57 You know,

1:08:57 we get to better entrepreneurs,

1:08:59 uh,

1:08:59 either from,

1:09:00 from better selection or from,

1:09:01 or from,

1:09:02 or from better training.

1:09:03 And then,

1:09:04 on administrative data,

1:09:05 so,

1:09:05 so funny,

1:09:06 I've spent most of my career kind of

1:09:08 justifying why survey data should be thought

1:09:10 of as as reliable as administrative data.

1:09:12 So it's,

1:09:13 so I,

1:09:13 I,

1:09:13 you know,

1:09:13 I'm,

1:09:14 I'm happy to hear,

1:09:15 uh,

1:09:15 people say that,

1:09:16 well,

1:09:16 maybe,

1:09:16 and,

1:09:16 and it,

1:09:17 and it's clearly gonna be the case when we think about

1:09:19 taxes and,

1:09:20 and,

1:09:20 and,

1:09:20 you know,

1:09:21 reasons that people would,

1:09:22 would,

1:09:23 would skew administrative data.

1:09:24 I,

1:09:25 I think I think that the data are valuable,

1:09:27 um,

1:09:27 and I think the research that's been,

1:09:29 that's been done when thinking about

1:09:31 firm to firm networks and how networks,

1:09:33 uh,

1:09:34 that,

1:09:34 that,

1:09:34 that,

1:09:34 the data that,

1:09:35 that,

1:09:36 um,

1:09:36 that,

1:09:36 that,

1:09:37 that the VAT data in particular allow us to,

1:09:40 um,

1:09:40 the,

1:09:41 the patterns of people that,

1:09:42 that those data allow us to understand is,

1:09:44 is,

1:09:44 is,

1:09:45 uh,

1:09:45 I'm gonna say unique cause it's really difficult.

1:09:48 It would be really difficult to,

1:09:49 to,

1:09:49 to,

1:09:50 uh,

1:09:50 uh,

1:09:51 generate any kind of broader,

1:09:53 um,

1:09:54 Uh,

1:09:54 survey.

1:09:55 Um,

1:09:56 and,

1:09:56 uh,

1:09:57 in,

1:09:57 in a broader sense,

1:09:58 in a,

1:09:58 in,

1:09:58 in a survey.

1:10:00 And,

1:10:00 um,

1:10:01 and then the question of,

1:10:03 of

1:10:04 representation of the countries where these data are available.

1:10:08 If I go back 10 years,

1:10:09 I'm,

1:10:09 I don't think VAT data were available in any country.

1:10:12 Now,

1:10:12 we've got,

1:10:13 you know,

1:10:13 4 or 5 countries in Africa where,

1:10:16 um,

1:10:17 uh,

1:10:17 you know,

1:10:17 where there are at least 3 countries where they're very broadly available and,

1:10:21 and,

1:10:21 and,

1:10:21 uh,

1:10:22 And Ethiopia as well,

1:10:24 where people are using them.

1:10:25 So,

1:10:26 I,

1:10:26 I think

1:10:27 that

1:10:28 it's likely that

1:10:30 perhaps by demonstrating that other countries will,

1:10:32 will be,

1:10:33 will begin to,

1:10:34 will,

1:10:34 will begin to see these,

1:10:35 these,

1:10:36 these,

1:10:36 these kinds of data pop up in other places.

1:10:38 So,

1:10:39 um,

1:10:39 I,

1:10:39 I,

1:10:40 I,

1:10:40 I

1:10:40 Take these,

1:10:41 I take these points,

1:10:42 but I,

1:10:42 but I think that um

1:10:44 there are a set of questions that I think it's gonna be quite challenging for us to,

1:10:49 to address with,

1:10:50 uh,

1:10:51 uh,

1:10:51 with,

1:10:52 uh,

1:10:52 with,

1:10:52 with survey data where I see uh promise in those,

1:10:55 uh,

1:10:56 in,

1:10:56 in those,

1:10:56 in those data.

1:10:57 And

1:10:58 thanks.

1:11:01 Thanks,

1:11:01 Chris.

1:11:02 Deborah.

1:11:04 On my,

1:11:05 on my side,

1:11:06 maybe,

1:11:06 um,

1:11:07 one of the things that I was uh trying to pass

1:11:10 as a message is also the fact that that combining data,

1:11:14 data,

1:11:14 uh,

1:11:15 it's,

1:11:15 uh,

1:11:16 extremely valuable.

1:11:17 On our side,

1:11:18 what we do is we combine the balance sheet information of the firms that are.

1:11:25 Uh,

1:11:25 externally,

1:11:26 so we have hard data combined to survey data

1:11:29 and then matching with,

1:11:31 uh,

1:11:31 all the data,

1:11:33 data sources,

1:11:34 databases on the location of the firms,

1:11:37 and,

1:11:37 uh,

1:11:38 all information that you have on the

1:11:40 overall external environment and I think it's uh

1:11:43 the combination of information.

1:11:45 So that brings a lot of value in understanding because then you can

1:11:51 understand how much is the firm behavior,

1:11:54 how much is the external environment,

1:11:56 how much is the policy intervention

1:11:59 that influence what the sing,

1:12:03 what motivates the action of the firm.

1:12:06 So I think that particularly in topics

1:12:10 extremely difficult to analyze like the the the green transition

1:12:15 where

1:12:15 data,

1:12:16 data are difficult,

1:12:17 the combination of

1:12:19 different

1:12:21 ways of seeing the same phenomenon,

1:12:24 including the survey data is important.

1:12:31 Thanks,

1:12:31 Deborah.

1:12:32 Uh,

1:12:32 so,

1:12:33 Dean,

1:12:33 over to you for the last word.

1:12:38 Sorry,

1:12:39 sorry.

1:12:40 Um,

1:12:41 uh,

1:12:41 yeah,

1:12:41 no,

1:12:41 I think this is,

1:12:43 you know,

1:12:43 for what it's worth,

1:12:44 I think this discussion on data has been fascinating,

1:12:46 and

1:12:47 I,

1:12:47 I,

1:12:48 um,

1:12:49 haven't worked as much with it personally,

1:12:51 with the kind of administrative data,

1:12:52 but I think it,

1:12:53 it hits on some of the,

1:12:54 I,

1:12:55 I,

1:12:55 I'm,

1:12:55 I'm excited by the,

1:12:56 by the promise of the integration,

1:12:58 both because of the scale,

1:12:59 administrative data possibilities and

1:13:02 And also because I think of the importance in trying to get at some of the,

1:13:06 the general equilibrium issues that I,

1:13:07 that I was talking about,

1:13:08 um,

1:13:09 that,

1:13:09 that is,

1:13:10 you know,

1:13:10 clearly a,

1:13:11 a path to trying to,

1:13:13 um,

1:13:13 to try to tackle some of those.

1:13:14 So it was,

1:13:15 uh,

1:13:16 thanks for having me on this and I was,

1:13:18 you know,

1:13:18 really was,

1:13:19 it was great listening to everybody and

1:13:21 learning more about what everybody's um up to and where

1:13:24 everybody is thinking there's big gaps and rooms for progress.

1:13:27 So I'm excited for the next 20 years.

1:13:33 Thanks very much,

1:13:34 Dean.

1:13:34 That's actually sort of a segue into what I wanted to say by

1:13:38 um

1:13:38 to,

1:13:39 to wrap up this.

1:13:39 So this is not just wrapping up this session in today's event,

1:13:42 but also the series of 3 events that we've put on to,

1:13:45 to celebrate these 20 years of KCP.

1:13:48 I keep focusing on the 20 years,

1:13:49 but I occasionally drop in and $80 million of over

1:13:53 $80 million of donor funded research uh during this time.

1:13:56 So,

1:13:56 this has really been a very significant effort on the part of our KCP donors,

1:14:01 which I think over the years we've had over

1:14:03 20 different donors to KCP who have contributed,

1:14:07 uh,

1:14:07 to make a lot of this research possible and to advance on,

1:14:10 on,

1:14:10 on this agenda.

1:14:12 I did want to

1:14:13 particularly acknowledge that our current donors to the,

1:14:16 the current round of KCP,

1:14:17 so,

1:14:18 uh,

1:14:18 CETA and Sweden.

1:14:19 Uh,

1:14:20 um,

1:14:20 France,

1:14:21 uh,

1:14:21 the government of Japan,

1:14:22 and also the EU through two different,

1:14:24 uh,

1:14:25 uh,

1:14:25 DGs that have been

1:14:26 contributing to,

1:14:27 uh,

1:14:28 to KCB funded research.

1:14:29 Uh,

1:14:30 um,

1:14:31 I think this is,

1:14:31 uh,

1:14:32 it's a great foundation for the next 20 years and

1:14:34 we're looking forward to being able to do much more.

1:14:37 Um,

1:14:37 just to

1:14:38 close up,

1:14:39 I also did want to give a big thank you to

1:14:41 Karina and Bintao for all the work that they've done in,

1:14:43 uh,

1:14:44 in putting together again,

1:14:45 not just today's event,

1:14:46 but the 3 events and for making KCP run.

1:14:49 Also,

1:14:50 this is now becoming more of an inside the bank comment for a second,

1:14:53 but

1:14:53 I did also want to,

1:14:54 uh,

1:14:55 thank Bintao in particular,

1:14:57 who's been with KCP for as long as I can remember.

1:15:00 Although my memory is failing these days,

1:15:02 so,

1:15:02 you know,

1:15:02 maybe that isn't such a big compliment,

1:15:04 but Bintao is going to be moving on shortly to,

1:15:07 uh,

1:15:07 do a new opportunity in DFI,

1:15:09 right?

1:15:10 No.

1:15:11 Water,

1:15:12 sorry,

1:15:12 in the water global practice

1:15:14 and sitting beside Bintao and just to introduce

1:15:16 quickly and we'll do it more systematically later,

1:15:18 but we have Anna Bokina who is joining us from the Africa region,

1:15:21 who will be stepping into Bintao's role.

1:15:23 So,

1:15:24 um,

1:15:25 welcome,

1:15:25 uh,

1:15:25 welcome Anna as well.

1:15:27 So with that,

1:15:28 um,

1:15:29 uh,

1:15:30 uh,

1:15:30 let me again thank our speakers,

1:15:32 David,

1:15:32 Jorge,

1:15:33 and our,

1:15:33 our panelists,

1:15:34 Deborah,

1:15:35 Dean,

1:15:35 and,

1:15:36 uh,

1:15:36 and Chris,

1:15:37 and,

1:15:37 uh,

1:15:37 you know,

1:15:38 big round of applause and everybody for all of their,

1:15:41 uh,

1:15:41 their hard work to make this happen and for all the interesting insights.

1:15:44 Thank you very much.

showAllTimestamps
no
transcript
I'm just going to uh turn directly to our, uh, two presenters. So we're gonna have the same format as the previous um session. So we've got, uh, Uh, two of our deck colleagues who are going to kick off the discussion with, uh, with complimentary presentations, and, uh, and then we'll have 3 panelists, and, uh, I had some inside information that unfortunately one of our panelists had a flight canceled and could not join us. So I only have 3, which buys us the 10 minutes that allows us to wrap up on time. So hopefully that will work out. Um, so let me, uh, sorry, who's going first, Jorge? OK, so, um, so we have, um, Jorge, not Jose, um, showing up, uh, to, uh, do the, uh, to, to, to kick off the discussion followed by, uh, uh, by David McKenzie. Um, so go ahead and I'll, uh, warn you around 12 or 13 minutes. Uh, OK, perfect. Thanks. Thank you, Art. Uh, just as a matter of organization, um, um, let me see, um. Do we have the But what this comes from, oh yeah, OK, um, uh, we, the way we organize the presentation in rather than being comprehensive, we actually selected a few topics that we think, uh, uh, there's data gaps and data and knowledge gaps, and we, in the process of presenting these different, different gaps, we try to highlight the contributions of KCP over the years. Uh, and I would like to start with a historical perspective of, of, of, of, of the, uh, dimension of the private sector, uh, development issues because, uh, the 20 years of the KCP, uh, match very well the development of the PhD agenda. The KCP started funding considerably, uh, considerable amounts of, of data collection efforts on firm level service, uh, which coupled with the WDR of 2005 on, on 2005 on investment climate and of course previous resource at the bank on. The binding contrast to private sector growth explain the generation of the creation of the two main diagnostics of the private sector, uh, that have been, um, the drivers in a sense of most of the operational policy in private sector development, enterprise service, and the doing business indicators, um. From here the link between research and um policy is in the cases of the in the case of the private sector development space mostly driven through uh the creation of these indicators that as we know are a major tool of our operational people to to in uh in here at the bank and our clients to galvanize all these principles of research into actual policy. Um, over the years, KCP has funded others, the, the research that, uh, substantiates other indicators like Finex, global Financial development in, in, uh, uh, development database, and for example, this issue, the, the, the, the childcare development or childcare services for the, uh, the. Important childcare development services have been demonstrated first in research and then had been made the case for incorporation into women's terms of the law. This is a very specific example of the of how the I see the the process of the linkage between research and a policy action has taken place in in PSD. Um, moving on to the, to highlight some of the issues that we wanted to highlight, uh, in, in the issue of informality, KCP has funding similar research, uh, on, on, on, on the, on the sector, and, and, and that is, uh, for example, I'm gonna draw again on the work of David McKenzie and Christopher Guro that are the panelists today on Sri Lanka in 2008 2010. 9 in this quite a few papers and the research papers that have produced by them, but in here I want to highlight the, the fact that uh this, this was, is a seminal because it, it funded the creation of all the, the, the whole generation of data from the, from the scratch, from starting from the full enumeration of all the informal businesses that operated in Sri Lanka, even in the cities of Sri Lanka where the study took place. Um, this is important because I want, what, uh, I wanted to highlight that because really, this really addressed one, the first, uh, data gap that I wanted to highlight on an informality, which is the lack of census data, of any data for that matter on informality in many countries. Uh, there's very limited information on the, on the extensive margin, the type of the businesses that operate informally and also in the intensive margin, uh, businesses that operate, uh, um, with informal practices or formal business operate with informal practices. Um, we have been working on the former, uh, at least in our, in our unit, with, with that incorporating, um, new techniques, geocoded data collection, coupled with a new sampling strategy that, uh, takes advantage of the clustering nature of these businesses in, in, in, in, in, in, in, in, in general in the world. And um by um this method, we are able to generate representative samples of the sector in a given region. That allow us to study whatever it was, you know, to do our, our, our, our, a questionnaire, to implement our questionnaire and at the same time, obtain an estimate of the population totals, in other words, the size of the sector. Uh, this is something that I'm here I'm showing, for example, as you can see in this, uh, for this is from the paper that we have written. Uh, we got hold of administrative data of census data for Ezotini. Uh, uh, that was fortunate because that allows us to actually evaluate the relative merits of this methodology vis a vis simple random sampling or certified random sampling, and in this paper, we show that there are efficiency gains in terms of the level of effort required to, and to implement these studies and not only that, we can, we're able to implement this in 3, in between 2 and 4 months in a, in a given region. Um, Moving on to more subject matters, uh, and to actually highlight a few of the knowledge gaps in informality as well that we find, I want to highlight three points. The first one, again, and I'm, I'm drawing from the same work of David McKenzie, this is, uh, in fact, this slide is David's, uh, but in this one is the, the lack of evidence of the benefits of formalization and the reduction of, of cost of registration. Uh, this has been shown by them and many of, uh, several other, other, other studies in here. This is again with the Sri Lanka study, as you can see on the left. There's a sharply decreasing demand for formalization and this is backed out for an experiment in which, er, informal entrepreneurs. were offered free registration in randomly assigned monthly profits. And as you can see on the right-hand side, the cumulative distribution of profits of the control and the treatment is basically, pretty much the same. They look alike. It is interesting that in this result, that uh the lack of evidence of uh uh uh benefits of formalization in 2008, 2009 in Sri Lanka, the authors mentioned that, uh, uh, by interviewing these, these entrepreneurs, most of them expressed that they actually were aware of the, the process of formalization. They just wouldn't want to know where are they taking it. We, I would like to compare this to our recent work in Africa, using that methodology that I just mentioned that allow us to do a representative sample of the sectors in several countries. In the, in the context of Africa, most of these entrepreneurs actually do respond that they do not know anything much about formalization. It's something that is not in the, in the greater skin as an option. Um, two other points on the data and knowledge gaps that I would like to highlight is the, the one on the left is something that has been shown by many studies already, uh, which is the, uh, the lower levels of productivity of those informal businesses when compared to similar microfirms. As you can see here, these results come from a paper written so, with the, with the data that we have recently collected. Uh, and, uh, the informal firms are, uh, uh, as you can see in, in, in, in blue, and the, the micro firms are firms that are very much like the informal firms in terms of sector size and even the business environment they face. Uh, the only differences will be fine between them is in the management practices, education of the entrepreneur, and the, uh, uh, experience. Uh, we think we still have to do a little bit more digging, more research to see these are the, the real source of such differences in productivity of the rest of, are there other things at play. A third point as well on the, on the knowledge gaps that I would like to highlight in the, in, in the, in, is the effect of the presence of informality on the formal sector. This is, uh, on the right-hand side, the panel comes from a paper also produced here from Amin and others, uh, in which, with more than 100 countries around the world, and as you can see, there's a clear difference between the firms that face competition from formal firms, the frames competition for informal firms for, and those ones that do not face competition for informal firms. This result is a result that other Papers, uh, other, other authors are also found with all the different uh metrics, uh, innovation, TFP in a different direction. So, I don't wanna make too much about the difference here in labor productivity, but more of the mechanisms at play, er, that we need to research a little bit more because on the one hand, uh, one would think that, uh, the presence of informality is more competition for formal. Firms and logic would say this firms will become more effective and more innovate more, and more productive. Uh, but on the other hand, this access, these informal firms have access, uh, that is in a sense, preferential or unequal, unfair competition, and therefore, they, they have access to, uh, inputs and services, uh, at lower cost, sometimes even free, illegally. Um, let me just move back quickly to the formal, to the formal sector because I, I don't really want to give the impression that by all the diagnostics on the first slide and, and the, and the, all the operational work that has derived from that, that the, in the formal sector, there are no data gaps. And I think one, in the presenting here, one, and I'm less optimistic that we will be doing, we will be able to do uh something. effectively in the medium term, which is the lack of administrative data to produce research. And I'm using here the a KCP funded project from the Mary Howard and others did in 2013 in which they try to collect uh these indicators on firm dynamics, productivity, and, and, and, and resource allocation uh from administrative data. 10 years later, now 223, it's only a handful of countries where we have administrative data and the issues, uh, the obstacles that we face in accessing, uh, administrative data, uh, are, are relatively serious in my view because there are legal issues on how the data was collected, uh, and, and difficult to overcome. And when one overcome those, those issues, these issues of comparability and quality. So, I think this is an issue that will remain er for, for some time. And if you follow my premise here, I think the usefulness of the relevance of survey data becomes even more relevant, more important. Uh, and so I want to highlight two points I saw as well that we, we think there is some, uh, knowledge gaps and data gaps in survey data. One is related to a point that was being discussed earlier, is the need to have more data, uh, tailored for services and to understand, well, to, to understand better multi-factor productivity for manufacturing as well. But mostly for services and the second point that I would like to highlight is the need to have more um frequent and uh uh frequent and uh a larger panel datasets that will let us address the issues, the issues of causation and move away from the issues of correlations. Uh, I don't wanna over, overstress this because, uh, KCP has funded quite a bit of research, uh, that, uh, on this and a lot of cross-section data that has been very illuminating, uh, on the evaluating, for example, the efficiency of the state-provided services for the private sector. Here, I have one paper that I find very, very. Relevant, which is by, by Caroline Freund and Bod Rikers and others here in DC, in which they have shown that, uh, and tested that the hypothesis of greasing the wheel of bribery, that it is, it, that it, that it does not hold. As you can see in this example here is one of the many examples of different transactions that firms interact with the government. Um, let me just finish here, uh, uh, that with one final point that I would like to, uh, highlight, which I think is an area in which we will need in thinking on what I was saying earlier on thinking on to the future, uh, which we, we, and I think this area is one because we need, it's interesting and relevant. Because not only is it important, but we need to think not only on methodologies, how to address it, but also new data that we need to be collected. And, and this is the, the economic cause of discrimination along the lines of gender, race, and even sexual identity. These are populations that are difficult to, to, to, to collect data from, uh, and, uh, and there is not quite clear even how to go about measuring this economic cost. That's what I'm using here. Uh, this very relevant paper and I find very impressive, uh, from in Econometrica in 2019, which does this, at least for the US and as you can see, it compares the GDP growth of the, the actual GDP growth in blue with the GDP growth hypothetical that will emerge from their model, uh, if the restrictions, uh, the discriminatory practices of black men and women in the US in the 1960s will have prevailed. I can, you can see that it's clear in the graph, the, the big difference between the two, which in the paper they say is about, uh, accounts for about 40% of the GDP growth uh between 1960 and 2010, which I think is, it makes the case for how important this is. And with that, They did Great, thanks. So, um, I'm gonna dig in on sort of one topic that the KCP has helped illustrate this sort of going back and forth between data, research, policy, and, and back to more data and more, more research. And so, uh, this is the sort of, Question of, you know, why the firm size distribution looks the way it does in most developing countries. And so, you know, if we look, most firms in developing countries are very small, many are informal, their productivity is very low, uh, and so, you know, this raises three questions about, you know, why are there so many small and unproductive firms? How much of a problem is, is this, uh, and, and their informality, and then, you know, what can policy do about it? And so, you know, if we want a framework for thinking about, you know, the firm's size, then we can sort of think about the problem of a typical firm owner who's trying to choose their inputs of capital and labor to maximize their profits, and they have some sort of production function, which has got this, um, capital and labor that they're buying, and then this term theta that they're, um, using that, that helps them turn their capital and labor into, to output. And so this suggests, you know, 3 potential explanations for firm size, um, being the way it is. So one is that firms are small because they're constrained, and we have, you know, a whole other programs on finance that we're, we're not gonna talk about today or on labor markets that we're not gonna talk about today, but thinking about sort of constraints for firms and getting the inputs that they want. The second that I'm gonna really dig into is that firms are small because they have, uh, a low theta. They're just not very good at, um, turning their inputs into output, and so we're gonna dig into that. And then the, the third point, which is, um, sort of related to the informality point that Jorge was talking about is that, you know, firms are small because the regulations are affecting their choice of, you know, their production function here, should they be formal or informal, and, and that can also affect the cost of their inputs. But I'm gonna dig into this sort of middle explanation, which is, Um, you know, maybe firms are so small, because they're just not very good at, at managing their, um, process of converting inputs into outputs. And so, uh, you know, how do we go about measuring and improving this, this, uh, term here? Well, you know, where were we 20 years ago? Sort of 20 years ago, there was this recognition that there was something that we called managerial technology that affected the ability of firms to translate these inputs into outputs. It's, but it was really a black box, um, so, uh, you know, there's famous papers that sort of say, whatever this is, it's, you know, some firms are better at it than others, um, or there was this sort of view that management was really, you know, uh, to do with style, or with charisma, or, you know, something that was, was hard to potentially teach, and some people were just born with it, and others, um, were, were not. Now, you know, of course, there was a, a wide range of business training programs that were based on the idea that there were some things that you could teach, um, but there weren't really a lot of evaluations or no evaluations of these programs, and there was less policy attention to what could be done in larger firms to improve this as well. So, uh, this is where the sort of first step of, of really just trying to measure and understand, um, comes through. And so, uh, you know, there was a set of influential, um, surveys done in large firms by Nick Bloom and John Van Reenen that, you know, really shaped the idea that these, That management was, was not just a, a style or something that you're born with, but it was really a set of concrete practices about how do you manage good workers and bad workers, how do you deal with quality, how do you plan for the future, things like that. And in the KCB supported work, uh, Chris Woodruff, who's, who's gonna be one of the panelists, and I thought about how do we, Take those same ideas and, and measure this in small firms, and do, do these things matter in small firms? If you walk into a lot of small firms, they're not keeping records, they're not doing any marketing, they, they don't have a budget, does this matter? And so what we see is, you know, across a range of countries, both in the cross section over time, it seems like, Firms differ a lot in these practices, that firms that do more of these practices are more profitable, sell more, and um grow faster. And so it seems like there's something about, you know, business practices that would be good to try and get firms to, to improve. Uh, this is also something that the, um, enterprise surveys unit has been building on with, with, um, adopting, how do you take these, um, ways of measuring management that were done, sort of, uh, um, in this sort of very expensive double-blind system and adopt them to how they can be done in firm surveys. And so, you know, they've been adopting both of these large and small firm, um, techniques to now make this available for a lot more countries. So then, you know, once we have this idea that, that there's something about these practices that, that seem to matter, the question is, you know, can we do something to improve them, and, you know, does this, this matter? And so this is where I wanna sort of give an illustration of this interplay between, You know, the KCP supporting what I call blue skies research, where it's not really testing a policy that we think about testing, it's not linked closely necessarily to a particular country operation that we have at the moment, but it's trying to generate, you know, where should be things, things be going in the future. Um, you know, is, is there even a, should we even be trying to push policy in this direction? And so the KCP supported this original work that, uh, I did in India with, with Nick Bloom and, and others on sort of a proof of concept, which was, if you take a badly managed firm, can you actually go and improve management in those firms, and would it actually matter? And so, you know, this was definitely not something that we'd recommend for policy. We hired the most expensive consultants you could find on the planet, probably. Um, uh, you know, very expensive global consulting firm came in, did really intensive consulting on these firms, uh, and, and, you know, spent months in these firms and improved management practices. But the key thing here was that we could see, once you improve management practices, we could track in these data that you could get quality improvements and see total factor productivity improve. And so, We could, you know, show not just that there was this association between better management and better practices, but that there was a causal link. If you could improve management, then you could get this, this increase in productivity. So this was, you know, nice, but this was not, you know, where you would necessarily want to go with, with policy, doing the exact same thing. And so, you know, then, um, but this is how this sort of research can then help generate new operational ideas. And so my operational colleagues, you know, are talking with finance ministers, productivity is always mentioned as an issue. And the question is, you know, what have you got that, that tells us how we could improve productivity? And they said, well, you know, there's this work from India that shows that management can, can help. Um, and, and, you know, is there a way We could think about doing that in Colombia. And so then I worked with them in, in, in the government in Colombia, and we sort of tested, could you do this, but could you do this in a more scalable, more cost-effective way? And we tested a group-based consulting, uh, method, as well as this individual consulting method. And you see in this graph, you know, across a whole range of management practices, you get improvements, um, that are very similar from this group-based consulting approach than you. from, from the individual approach at a third of the cost. So, you know, we get this, this method that can improve, um, some growth in Colombia, can be done at a lower cost, and, you know, then that link, links to, sort of, the next round of policy where now, now that that's been done in Colombia, countries are coming and saying, oh yeah, let's see if we can do this in Ghana, we can do this in Malawi, etc. And so, um, You know, that's, that's sort of, you know, from that, that blue skies research leading to new government projects. And then, you know, one of the things that comes up is that there's just not many of these firms. We're trying to do this a lot with sort of small samples. And so, one of the things that the KCP has also been, um, helping us do is more methodological improvements on, you know, how do we incorporate prior information, how do we make the most of these small samples. And so some of the most recent work we've been doing on uh new Bayesian impact evaluation methods has been supported by the KCP there. So then, you know, that's large firms. What about small firms? Well, you know, some of, uh, one of our panelists, Dean Kalin was, you know, really one of the first, I think the first to, to look at, um, was this business training effective. And there's been sort of many studies since that have found that, you know, some of these business training programs can help firms, even the very smallest firms, improve their business practices, improve profits and sales. But, um, you know, it's, it's somewhat mixed evidence. Uh, in some cases, it's not always as factors as, as we'd like. And it seems quite hard to scale some of these in-person classroom-based, um, training. And so, that's where, you know, I think two of the other directions that have been coming through some of the most recent KCP supported work, I've been doing, my colleagues, Miriam and, and Kayle here have been working in Brazil to see if you could sort of start with a much lighter touch approach, give some firms diagnostics, um, use this to identify, you know, which firms you might then want to support in the second stage and get firms to demand some of the training themselves. Um, unfortunately, our, our panelist, um, Geo, who, who's, uh, on a flight right now, uh, she heads this, uh, organization of, uh, female entrepreneur, um, training. Programs in, in Mexico called CREEA, and, you know, we've been working with them to see if you could do, use digital technology to try and scale um training and, and, you know, provide it remotely, and we've been working, you know, both in Mexico and Guatemala, you know, using Zoom, use, using what we're doing right now to do training to, to people all across the country. So, you know, we're, we're trying to learn that. So, just, to sort of, you know, where, where do we go next? I guess, you know, there's a couple of recurring themes of persistent puzzles. And so, one is just, you know, firms are just very different from one another. Management capacity varies dramatically across, um, and within countries. There's just lots of variation in the scope for growth. And so the question is, you know, how can we identify which firms we should be targeting, um, policies to, um, more? How can we scale this in a way that, um, You know, the market works better, how much does the government need to be involved versus just trying to get the market to work, um, better, and then we need to, you know, work on collecting better, um, data on firm, firm outcomes, which is quite tricky. And so, you know, as a, uh, you know, then there's these new areas of private sector development that are really coming to the fore in terms of, you know, how do we collect more information on, on green growth, green PSD, um, transactional government, the enterprise surveys, you know, have been starting to measure these practices. And, you know, hopefully, this is the next phase of, you know, we collect some data, we identify some diagnostics, and now we can, you know, start some interventions based on, on that. And so, um, you, you know, there's, there's just sort of a lot of scope now with big data, with AI with other approaches to try and improve our, our survey technology, and I hope this is sort of gonna be, you know, we start a whole new batch of, uh, diagnostics that can help get this cycle going again. So let me stop there. Thanks very much, Jorge and David, um, uh, for great presentations, uh, or presentation, I guess very integrated, um, so we'll turn now to our panelists and as they've already mentioned, uh, unfortunately, uh, Giovanna Hernandez Constantino, um, who was supposed to be our first panelist is unable to, to join us because of, uh, travel mishaps, um. Uh, but we do have, um, uh, 3 other excellent panelists who we'll go to in turn in the same format as before. So, uh, we'll begin with Dean Carlin, who, uh, needs a little introduction, um, and, uh, currently is the chief economist at USAID. Um, then we'll turn to Deborah Revoltella, who is the chief economist and director of the Economics Department at the European Investment Bank. Um, and finally, we'll turn to, uh, Chris Woodruff, who is at Oxford and, uh, uh, also I think needs a little introduction given the, uh, the importance of the work that he has, uh, uh, he has done and that we frequently see cited here in the bank. Um, so, starting with you, Dean, uh, we'll, uh, we'll go with, uh, my, sorry, I think the commitment was to 8 minutes each. So, if you can sort of aim for that and, uh, Uh, and, and then we'll have a chance for a discussion. Uh, so over to you, Dean. OK, hold on, I'm starting a timer. Don't start the 8 minutes yet. There we go. Um, hi, everyone. I, I wish I was over there. Um, sorry, I, I ended up too tight, um, on the, on the, on the walkover. So, I'm going, I'm doing the lazy path. Um, so, I wanted to share 33 thoughts that, um, that also feed, you know, build on a lot of the things that we just discussed, um, and, and, and are by no means in lieu of, I think there's a lot of, like, really important questions that, um, that, that we just heard about that are being posed. There's a lot of different directions where more research is needed to understand. You know, things like the technology question that David was just talking about, um, for, for achieving scale and, um, and, you know, we, we have an overall I don't want to say conundrum, but a little bit of a conundrum of the fact that, you know, I think, I think we can kind of go backwards and say, if someone asks you for a point estimate, You can find the study that, that finds it when it comes to things that are in the space of promoting managerial capital or business training. And, um, and that shouldn't be seen as a, as a, in some, in some respects, I think that should be seen as a motivation, not as a, not as a deterrent, a motivation for, for the kind of work that we've just heard about, that's really trying to get in more granular understanding of Of both, how do we do, uh, diagnostics as well as how do we, how do we separate out and understand what types of implementation strategies are, are, are working and the intersection there and hopefully we'll start, um, getting more and more insights. But as, as you heard, there's, you know, there's definitely You know, we, we, you know, the state of the evidence is definitely better than um we can, that everything goes, right? It's, it, we, we do, there are some clear patterns that are coming through that, that show um policymakers who are not interested in research, you just want to use evidence that does show we, you know, we do have important insights as you've just been hearing about. Um, so I want to highlight three areas that I see as kind of big gaps that in this, in this area that, and when I say big gaps, I want to be clear that like, there's definitely like really good, smart, um, careful, um, policymakers into, you know, implementers and researchers working on these, so, you know, To a lot of people in the room, um, there's a none of these three are gonna be like, wow, we never thought of that before. But, um, but I do see these as three areas that are continue to be, um, areas of big, big gaps relative to the kind of standard, um, impact evaluation. The first is very much motivated by my baldness. Um, And um, although the jokes don't seem to work too well over Zoom, but um I have a real gripe with hair, um, and, and with hair salons. And it's just to make a very simple point about general equilibrium effects that we don't, you know, we talk about them a lot. We obviously thought about them a lot. There's been a lot of good research on it and, you know, by, by people, by David and others, and, um, and, but, you know, obviously, no pun intended, there's not gonna be a very simple generalized answer to generalized equilibrium, general equilibrium effects. Um, and yet we desperately need to understand more about them. Um, there's been, you know, a lot of, when we look at the expansion of microenterprise in the US, I was always struck by how often the prototypical enterprise that, that was being promoted as a microenterprise in the United States was hair salons. And this is why I make the hair joke, because like, adding more hair salons in, you know, in most communities, uh, you know, there's only a fixed amount of hair in the world. And, um, as much as I might resent those people, um, they're adding more hair salons does not lead to more haircuts. Um, and so, uh, you know, what, what's really happening here, even if you set up a nice randomized trial, you're very rarely gonna fit, you know, kind of detect that kind of effect. It's, you need a very large study, um, you need a lot of the stars to be aligned. But we do need more work to understand this. Um, and we need to first separate out and not, and not do what I just did and talk about general, general equilibrium effects as some sort of monolithic thing and, and separate it out into information effects, um, and price effects and then kind of industrial organization, competition structure effects. And the, so that's the first thing we need to do when we think about G and then, you know, could also think about kind of macro multiplier kind of um effects as well. So, you know, we need to separate out what we mean by this. We need to do, we need to try to incorporate when we can better, better disentangling of these effects. And then we, you know, we need to, we need to kind of rope in our macro friends as, as micro people and um to, to um integrate some of the, the micro studies into better macro modeling of these kinds of issues to understand more about how to set policy. That um can hopefully use GE effects to improve impacts rather than, um, rather than, you know, the opposite, which is the fear in many cases that they're going to, um, diminish treatment effects and, and all we're doing is business stealing by promoting some businesses. Um, the second gap that I, that I, um, have been, um, kind of obsessed with for a while, and I think it's actually much less, much less research compared to like the general equilibrium is, um, a long time ago I wrote a, uh, a kind of a thought piece with Sendel Mono, and been to Anno called Three Anomalies. Um, and one of those anomalies was, why don't we see more people share businesses? Two-person, three-person equity firms, um, as the on-ramp. And, and it's really easy to think through examples and household enterprises where there's a clear economy of scale. Um, by having or economy of scope, depending on the business style, goat herding, piggeries, cow sheds, where a cow shed is a fixed cost and having multiple households keep their cows. There's lots of examples like this that we, that are easy to tell. And, you know, what's the obstacle here in understanding more about those, um, those challenges? Is it moral hazard? Is it training? Is it, um, is it trust? Is it just, um, is it, is it just wrong and the returns aren't that high? To some of these, but it does seem like on paper, some of these returns could be really, um, really remarkable. And, and we need to see more work on that. So, that's not quite on the, that's, I would put that in the small, but not the medium size. This is about the, that, that bridge from getting from micro to small firms. And the third is why, here's a puzzle that I've always been struck by is, given that we do see some um really strong impacts from, you often more tailored and customized managerial capital intervention. I'm thinking about things like the India study that David did, I'm thinking about the Mexico study that I did with Antoinette um Shore and Miriam Brune. Um, and, um, these showed really high returns. Um, we saw some pretty impressive improvement in, in labor force, in returns to, um, in kind of total factor productivity in, in, in Mexico. So how is it possible, and David's heard me ask this question before, so, how is it possible that these consulting firms can be so good. At giving out advice, and yet they're not. Like expanding amazingly well and getting really, you know, rich themselves. What, what's the, what's the problem in the consulting market, if this really is such a, a clear path to improving SME, um, total, you know, profits? Where's the issue? Is it risk? Is there heterogeneity and quality, and people don't know how to choose firms? Is it information? Um, is it, um, is it capital? Is it just a capital constraints or a contracting issue? And that's my 8 minutes. Um, you know, where's, what, where's the market failure rate in the market for consulting? Like, and, and I, I'm, I'm still, I, I, I was struck by this from the study that we did in Mexico. I've been struck by this by any, any study that shows big success for managerial capital interventions from a for-profit entity. That is not then expanding magnificently themselves. Like what's, what's the, what's the constraints in that market? So I'll close there. Thanks very much, Dean. Um, so you want to talk to more macroeconomists, I'll just remind you that macroeconomists mean something very different when they talk about haircuts. Uh, so just bear that in mind when you start those discussions. Um, all right, over to Deborah. Thank, thank you very much and thank you very much for inviting me for uh for uh this uh panel and uh this uh discussion and I'm also very sorry not to be with you today um in Washington but uh being only remotely, but for me it would have been a, a longer traveling coming there. So, um, I, I read the task, uh, for being in this panel in terms of, uh, suggesting topics that I think, uh, in terms of private sector development at this point in time, we should, uh, uh, really put in the agenda and also, uh, think where we need to invest. Much more uh on uh data and uh research and I think uh um where I'm coming from and uh probably is also for uh the institution I work with um one of the uh key concern uh in this moment that uh that we are looking at on the research side is trying to understand. Um, what motivate, uh, the transition, uh, the, the, the transformation of of firms of, uh, of companies in the net zero transition and, uh, addressing climate change. And I think on that point. Point of view, we really, we really have an issue in terms of data availability and understanding. We're investing much more in terms of data and the survey also can play a rule would be quite relevant. Um, where I'm coming from, uh, um, we, uh, we know that, uh, uh, that the transformation will, uh, will the net zero transition will lead to a complete change in relative prices, reallocation of resources, and economic power within and across sectors, and the, the incentive. to move as a company and to transform really change a lot depending on where you position it. If you hold assets that will become stranded assets, if you are in a good position for developing a new technologies or not, whether you have access to. Resources that are necessary for the green transition and also what is the policy framework around you that incentivize or disincentivize the transition. So to me what what would be a very useful. Processes to, to start looking more at uh what motivates firms to invest in the net zero transition, to transform and radically transform why some firms are lagging behind and others are. Accelerating the transition while some play into gaining the upside while others just play into amortizing the current assets before doing anything in terms of transformation. And the complexity of doing the work is really that there are so many elements at hands both on the regulatory point of side, on the firm specific point of file side, the counter-specific point that uh that actually you, you really need to, to go very granular in terms of um of uh uh of data and uh. And information and that's why I was thinking that working with a survey may be a useful compliment. And maybe if I still have one second after this long introduction to explain why I'm showing this to you, I will try to share a presentation where I will just show a couple of slides, um. And I think, uh, can you see it? Yes, it looks good. OK, and I think I have to, OK. Um, what I will show you something that, uh, we launched, uh, 7 years ago at the European Investment Bank is a survey of more than 12,000 firms at the European level, representative for single countries and with a benchmark at the European, uh, at the US level and also is the links to balance sheet data of the firms. So we combine the balance sheet information of the firms. With uh qualitative and quantitative data that uh we um we uh gain from uh from uh the survey exercise and what we can do is, uh, uh, first of all ask the firms to think about the barrier on investment here in particular in the NASA survey we had a very strong rule of energy costs as a barrier to investment and uncertainty at the European. Level and that's allowed us also combined to the other variables to understand what was happening in terms of climate investment and energy efficiency investment in Europe. So what we find out is exactly that in the moment in which in Europe we had the energy crisis last year coming after the war in Ukraine, what we had was that Energy costs were a driver for investment in energy efficiency, but on the other side, the uncertainty was was a drag to investment overall. But when we look at energy efficiency, the net effect of the push for the energy cost and the uncertainty left a small positive. So increasing the probability. Ability for firms to invest in energy efficiency and in fact we see it at the European level with very strong investment in energy efficiency this year in the private sector coming up. What happened with the climate action investment that is a wider concept not only energy efficiency but also adaptation and all other element of mitigation was in fact that the um Positive impact of the energy cost was a completely balanced by the fact of uncertainty. And we explain it also because on a policy point of view there were various instruments that were um Used also to contain energy prices to um to, to support the firms during the crisis that uh distorted incentive on investment. So what you had is that the combination of uncertainty and energy costs left a positive stimulus for energy efficiency but it didn't change the probability of firms to invest in climate action. And that's something important uh to understand also on a policy point of view, to know which kind of instrument and policies that work to stimulate the climate investment action. And then uh uh an example of what we are doing in the survey, we are asking firms what they do for mitigation and for adaptation and also their perception, of course, uh, before I don't show it of physical risk and transition risk. What you see on mitigation, we ask them which kind of measure they implement. You see a lot of use of energy efficiency investment of waste minimization that helps us understanding. The regulation of matters and actually in Europe we see the effect of a regulation in imposing a waste minimization for firms and that's why that's an area so developed. But it's also interesting that the um the investment in new less polluting technologies or activities which is the transformation of firms moving either changing technologies or moving in business areas and business activities. That are less polluting. What we see there and it's the current work that we are developing, we are trying to look at the most energy intensive sectors and looking at what motivates the firms that are actually investing to transform and those that are not transforming and with more binding regulation coming in, you would expect that those that are non-transforming, they are trying to use the assets. In the end, and then they will just close the activities while uh the others uh start transforming the business already now. I think uh further analysis and further work in this direction is very important because of the structural shift that the economy will have and understanding how to, um, how the reallocation of resources and also the transformation of firms that will happen is extremely important. With that, uh, I think, uh, uh, I had in the presentation a couple of additional slides to show a project that we developed actually also with uh Jorge and with DBRD based on the enterprise survey both on the North African region and on the East and North Africa second project. There as well we were looking at the physical climate risk, green investment and trying to understand that the uh drivers for firm transformation. And again I think that was an early um attempt with an early version of a green model for the enterprise survey. I think that going uh um. The more direct as possible, as direct as possible with the new set of definition that now start becoming uh um standard and understood also by the corporate sector in terms of what is the physical risk, what is the transitional risk, and what can be done in a mitigation and adaptation. Having a more spelled out question I would be. Useful and would help also uh around the world a better understanding of what motivates the film. I hope this uh um helps in the discussion and uh and uh that I didn't bring uh topics completely out of uh of uh what you wanted to discuss, but I think that the climate part is very important to consider. Thanks very much, Deborah. Um, uh, Chris, here, uh, our last panelist. Go ahead. Great. Thank you. And, and thanks for, uh, the chance to, to, to, to speak to you today. Um, first, I wanna say congratulations, uh, to, to KCP for, for the work you've done over the years and, and the amount of time. I think I wanna sort of both reflect back a bit over that period of time and what we've, I think, accomplished and also kind of think about what really is is maybe where we've done less, where we've been, where we've made less progress, and where I think the next 20 years, uh, might, uh, uh, might, might focus. And, and as I sort of reflected back at, at, at, on the 20 years, I, I thought about the work that David and I started doing, you know, I guess a bit more than 20 years ago. Um, and thinking about Um, data and microenterprises and where we were with kind of how we understood how, what profits were, how firms, um, thought about investments, thought about returns and things, uh, things of that sort in places where people, uh, didn't even, uh, keep, uh, any, any written records. Um, and the literature was really quite sparse in a sense. Um. I think one of the, if, if I, I think about that time, that one of the surprising um pieces of data that was out there was the, the, uh, the Enemin survey in, in Mexico, the microenterprise survey in Mexico, which, which David and I used early, early on, um, which was in some senses, way, way ahead of its time. And if I think about the 20 years and where we've come from, from that point in time, um, From the perspective of data and how we understand these firms, it's really quite amazing, uh, amazing progress. And, uh, And, uh, as, as was pointed out, a lot of this work, you know, uh, important work was funded by, uh, you know, by the bank and by, by KCP, um, but it, but we're, we're at a point where it's, we're way beyond kind of, uh, measuring, um, uh, what, what firm, how firms keep, how, how firms think about profits and, and, and, and, uh, and other things, and thinking about really using technology and using experiments and using, um, Uh, other kinds of, uh, information to think about, um. Uh, effort at, at, at, at their, when, when they're working in the enterprise and other aspects of, of, of productivity, um, that is, uh, is, is, you know, really quite some, some startling progress. And I, and I think the progress has been most, uh, uh, impressive and, and strongest in the area of, of microenterprises and thinking about these very small scale, uh, these very, very small scale, uh, enterprises. I think recently, there's also been a fair amount of work in very large enterprises. And, um, and in, in a way, these are the, you know, the, what's, what's missing is, is the missing middle. And it's not just cause the firms, there are not as many firms there, but because we, we haven't, I think, focused as, with some notable exceptions, haven't focused as much attention on, on sort of mid-sized firms. And It's kind of obvious as researchers why that's the case, that we end up, uh, going to the field and seeing massive numbers of very small-scale firms. It's easy to get a sample, we can do, we can do work, or we can go develop a relationship with a large-scale firm, and the firm's large enough that we can do experiments and other work, uh, within a single or a handful of, of, of very large firms. Um, but if I think about the kinds of firms that are, um, participating in incubators, accelerators, the sort of fast-growth, you know, sort of, uh, uh, uh, mid-sized firms, um, there's, I think, quite a lot, you know, less work in that area. And I think that's a place where, um, where we need, uh, we, you know, we need, uh, we need more attention. And I wanna, I, I guess I wanna say, I think, um, and, and Jorge, you, you mentioned the difficulty with, with administrative data and the lack of administrative data, and I think that's right. There's often in many countries, a lack of administrative data, but we're beginning to see now in, in some quite, you know, in some quite important instances of people getting access to value-added transaction tax data, to trade data, to uh corporate income tax data, uh, and being able to sort of merge across data sets uh in India, and Pakistan, and Uganda, and Kenya, and other, in other countries. And those Data, I think, have uh, a tremendous amount of promise in helping us understand um both firm dynamics generally and the sorts of general equilibrium effects that, that, uh, that, that Dean, that Dean was, was talking about. Um, they have the advantage of being comprehensive. They have the advantage of being, um, Uh, you know, available over a long period of time. Um, of course, only for formal, for formal firms. But, uh, but the, these, you know, these sort of faster growing mid-sized firms are, are likely, uh, to be. And so, I think the The one of the challenges we face as, as researchers and, and, and, and when, when working with, with policymakers is figuring out ways to, to make those, uh, those kinds of data more accessible. And, uh, for example, Uganda is setting up a secure data center. South Africa has, has done some work in setting up secure data centers to make the data more broadly available. I mean, I think over The past, they've been typically available when you get the right relationship and you some, and you, and you get access to the, to the data. But, uh, but I think the challenge of making those more systematically available is one that I think the bank can, uh, can, can play a, can play a role in, and I think would have, uh, uh, if you build it, they will, they will come kind of sense of generating a lot of research, uh, uh, a lot of research in this, um, in, in this area. And Uh, one of the things that I would say, I again see as an area that's developing now, and an area that's likely to grow is exactly on the sort of, uh, um, interaction between, uh, microeconomists and macroeconomists, um, or at least the interest in macroeconomists of, uh, in, in the private sector and in firms. Um. I run a, uh, some of you will know, I run a program funded by uh FCDO, uh, uh, called Private Enterprise Development in Low-income Countries, a grants program. There's a sister program called Structural Transition, uh, uh, and Economic Growth, uh, which is basically a macro program. And SEG funds really quite a substantial amount of work on firms and work on the private sector, and much of it is exactly in this area of kind of thinking about, um, Uh, you know, how we think about general equilibrium effects, how we think about using, uh, structural models, um, uh, to, uh, to, to, to, to come up with sort of policy counterfactuals. We're not quite there yet in, in marrying the, the microeconomics with the, with the macroeconomics, but, but there have been Really important moves in the, in that direction in the, in the last, just, you know, recent years. And I think that's also a place where, um, where there's a lot of potential, uh, in the, in, in the next, uh, in the next, in the next 20 years. So, I'll, I'll stop, I'll leave it at that and, and say, I think, you know, Measurement data, uh, and David mentioned the management practices. Another area where, um, uh, you know, uh, as, as the saying in management goes, what gets measured gets managed. You know, when, when we're able to measure things within, within, uh, Uh, within these firms and across these firms, um, you know, we were able to, to make much more progress in, in thinking about what policies are and what, what interventions are that, that will, uh, that will, will benefit firms and, and, and firm growth. So, Data measurement, yes, the way we've been doing it with surveys and with, and with developing kind of new, um, relatively small sample, uh, rich data sets. But I think also data, uh, thinking more about, uh, the administrative data and the, and the, the ability to kind of Reach much larger populations of firms, uh, where firms that are less common when we go to the field will, will, will be, uh, available and, and, uh, and there in, in, in larger numbers, uh, uh, will be, uh, will be important as we go, as we go forward. And I'll, I'll leave it there. Thanks very much, Chris, and uh I'm not sure if you're able to connect to the first part of the session, but you're really kind of bringing us full circle because we had a lot of discussion about the importance of, of drawing on administrative data, um, you know, both on firms, on tax records, and so on as a way of, uh, of generating insights and also the, the difficulties in accessing this data. So we have, excuse me, we have about 20 minutes um remaining uh for questions. Um, so the floor is open, we'll start with some in the room, and if we have any online, please, please let me know, Karina. So, uh, Sergio and then Daria, and then Karina if we have some people online. For us. And, uh, thanks for the presentation. It's very interesting. Um, one thing for advertising and, um, following what Chris mentioned is at the macro group and also using KCP funds. We are trying to link the micro and the macro aspects of the economy using, um, more widely available data at the country level or cross country. But, uh, my question was, uh, basically to David and, and Dean, you mentioned, uh, this issue of a black box of what entrepreneurship, uh, abilities are. Um, and you mentioned the value of, uh, doing consulting and training, uh, to these entrepreneurs that can increase performance over time. But it's still, for the audience, that's a black box. So, I was wondering whether you can uh provide a bit more examples of what are the practices that you use to train these entrepreneurs and what are the most effective uh training practices that you have uh seen that have an effect on, on, on entrepreneurial activity later on. Um, yes, my question, um, I think really to whoever wants to pick it up, but, uh, so there is, there's been a lot of discussion about what happens within the firm, but from the trade side, we see that really this firm to firm linkages are very important. So I was wondering whether there is, uh, um, you know, a program there to perhaps bridge that dimension. Now, I wanted to, I mean, it's maybe more common than the question, but um I think, you know, the, the link between blue sky type of, of research and then policy impact as, as uh David really, uh, you know, outlined very well and uh documented well, is really, you know, I think when you look at David's presentation, he really schemes uh through all the important steps in between getting that first, you know, uh, paper that was academically received. But then making those linkages with policymakers and really, you know, it's not like you have a paper in the top five journal and suddenly people hear your messages in the circles of influence, right? So, I think kind of thinking about together about the, the importance of KCP financing and the work of the bank more generally in making these linkages and, and can we think a little bit differently about KCP financing and making this work a little bit easier for researchers to really uh make that link. So, there's a question from Alvaro Gonzalez, which is actually similar to what Sergio just asked, and how can we focus more on the firm, the kind of firm dynamics that can promote better uh managed firms, and how do we better identify factors that really improve the quality of management. This is more for David. Thanks. Anybody else in the room wanna come in? Norman, please. Thank you. Uh, great presentation, David, Jorge. Uh, I want to make a comment on, on administrative data, uh, that has been the subject of the, of this session and the previous one. And, uh, present maybe a counterpoint on, uh, on the use of administrative data. I, I do wonder in some, for some countries, whether this data are truly reliable. Um, and whether we should be concerned about that reliability. Uh, and if that is a concern, then how can we Combine administrative data with independently collected data like we do in the, the Enterprise survey program to uh improve the reliability of such data. And the second point that is related is, uh, well, some countries are making this data available, but then, can we extrapolate to other countries that don't make the data available for some reason. Um, and my conjecture is that extrapolation is very hard. And that's why we also need independent data collection. So, 2 issues of administrative data. Thank you. Oh, thank you very much, Norman. I was actually going to raise uh perhaps even a sharper version of that question, because it's not just a question of reliability, it's also a question of, in some cases, political influence, agendas, and so on, uh, that we have to find ways to filter through with administrative data, and I, I think, well, particularly in your be Ready project, you're really trying to navigate that, um. Uh, let, let me add, uh, two other quick questions to the mix, and then we'll go back to the, uh, the speakers and the panelists for, you know, 1 or 2 minutes each for, uh, for responses. Um, uh, I mean, I'm, I'm picking up a little bit on sort of the, the similar question that, that Dean posed. Um, uh, so Dean asked the question about, you know, why is it that consulting firms aren't getting rich? And I think the flip side of that question is, You know, why isn't management sort of like just another input like capital and labor, and why are the sort of puzzles different in, in that case? Like, you know, for example, one could take the view that firms don't invest in management because they're small, and one might even make a case that there's some kind of a, you know, size trap mechanism that's driving that. Although I know from a paper that you and I wrote together, David, that we're not very sympathetic to that view. So, if it's not that trap view, uh, what would the other view be? Um, and I have a question for, for Deborah as well. So you showed some very interesting stuff on the green side, and I think you made a very compelling case that thinking about green incentives for firms are important. One question I had seeing the slides from your EIBIS survey is, you know, what do we learn from these questions about sort of private incentives versus external benefits of investments that firms are making, right? Because the trite answer always when we sort of confront that question and say, Well, your governments should just put carbon taxes and, right? But, but we know that governments can't put carbon taxes. So, we really need to understand, well, in the absence of carbon taxes, what are, what are the things that are actually in firms' private interest to do and what are the things that they're doing insufficiently because of externalities, and how can these surveys tell us more about that? So, with that, let's go first to Jorge and David, um, and then we'll go to the panelists, and if I could ask you to be a bit disciplined about sticking to, you know, 2 minutes each, we'll be able to wrap up on time. Yes, a few of the topics perhaps, uh, addressing first the, the issue of the um initiative data uh for uh Christopher, I think he raised that and Norman's point. Uh, the, the, the, I think I agree with Christopher in the sense that, uh, yes, you do, we do have a big gain access to some very good data in some specific countries. The point I was making, uh, is actually to the point he said that word systematic. Around the globe. It's, I don't think this is a very promising. It's, it's kind of like Nietzsche, what we can find in some certain countries, great, but when we try to think of that around the world, it's very limited. Uh, and then there's restrictions. I am interested in that among the countries that you mentioned, one of those is one of the countries in which I have my faced most difficulties accessing any data for, for any kind of work, for research and, and, and, and. And even for the implementary services in the case, the case of South Africa. Um, and I do know that we gain access to that access just to use as an example, we have gained access to that, but a very restrictive way of using the data and the bank has, has profited from that. But the point I was making is that it's a little bit uh difficult to put your money into that, into, into, uh, on a global basis and I realized we, the World Bank probably should play a role on that. Uh, uh, and, uh, it, it, it has come to me several times, the request of me playing the role in that way is because I interact with a lot of these, uh, with all the countries around the world in my role as manager of interpersonallysis unit, and at least a little bit of my frustration that I was expressing is, uh, my experience from, from building from all those interactions in which is, uh, is, is, is, it's, it's not, it, it, you just, we just, um, bump into issues of legality and it's not nothing researchy, it's poli, it's purely political, the access to this administrative data. Um. I think, uh, on, on, on, on, I just wanted to touch base on, on, on Deborah's point as well on the green practices in the green of firms in, uh, in the, the, the, she made the case of the work we did together in the past and, uh, uh, but she was making the case of uh trying to build more into the service, uh, this transition of the motivation of the firms into the transition. I fully agree and I would be happy to explore how they're doing it and see how we explore, but I also So, I would add, moving to the table that I would like to discuss how, what is the best way to, way to do that? Is it survey data, the best way to address that or is there other methods of data collection that probably would be better. I'm not very keen on asking direct questions about incentives to firms because I'm more keen on observing how they, how they act in trying to do that because um this, those direct questions on, on or perceptions in a sense, uh can be misleading. Um, I think just a little bit on, on, I think, uh, on the, on the, within the firm. There's a little bit of work on, on my, on my team, more on the research side, not on data collection within the firm on that, on what's happening within the firm, uh, but I, but it's true. It's not, not much systematic in the, at least on the data collection effort that we do have. It's a little more on the, on the, on the research side of, of a couple of people in my team. So, um, yeah, thanks, thanks to all these great, um, comments and, and discussing comments. Uh, so let me just pick up on a couple of things, and so, one is this, um, point we sort of get all the time, which is, We, you know, we work for years, we, we do a big sample, we, we painstakingly look at this data, and if we, and, you know, a lot of the times we find things that don't work, and then when we find something that works, The question is always, well, if it works, why isn't the private sector already doing this already, and, you know, I had to, you know, interview 3000 firms, trace them over, you know, 3 years, um, do a lot of econometrics, and I can barely detect the impact, and then we think that everybody should know that this is, you know, what, what's going on. And so, um, you know, I think it, you know, a lot of this stuff, it's, it's actually, even when you've gone through this, there's just so much other stuff that hits firms that it's actually very hard. Hard for them to know whether, um, what they're doing works or, or not. And so, this is, I think, one of the big failures in the consulting market is, is that it's, um, you know, it's an experience, good. You don't know what you're getting before you go in, as, as Sergio said, you know, you told me about consulting, but I don't really know what that, that is. Um, you know, there's a sort of set of practices, like, you know, even at a general level, if I say, well, it's about quality improvement and logistics improvement and HR practices, well, that's great, but then what specific practices, and, you know, that's all very context-specific. Um, so, I think it's, you know, very hard for firms to know what they're getting ex ante. We do see that when firms have gone through this, they're more likely to go back and buy a bit of this consulting on the market themselves. So I think there is this sort of experience, good aspect. But I think it's just really hard, um, you know, there's this great paper I like on advertising. Where, um, firms that are doing experiments with 2 million customers can't tell whether they're getting 50% return or 0% return on that advertising expenditure. It's just like super hard to know when so much other stuff hits your firm, whether anything you do, do works or not. And so, I think this is like, you know, one of the big challenges for us. And so that's where, you know, the, the sort of macro factors of, um, You know, what's driving firm dynamics and, and, you know, what else can we have at that sort of ecosystem level in terms of competition policy, and some of our colleagues have been working on that, and, and, uh, um, you know, what can large firms do through their value chains of trying to help their, their customers improve management and, you know, we see trade and having this role in improving. Um, and upgrading as well. And so I think it's, you know, trying to look from the bottom up and, and the top down, but I think the sort of point of, you know, when you finally find something that maybe works, um, why has everyone not done it already? Uh, and yet, you know, I, I do a lot, we also do lots of well-intended things that don't work, um, is, is there. So let me stop there and I'll, um, you know, I'm sure there's other questions, but I wanna give the panelists time to respond. OK, thanks. Let's go to the panelist's kind of in reverse order. So, um, uh, Chris, do you have a couple of observations? Yeah, sure. So, so a couple of things. Uh, one is, one is on the question of, um, on what works in training and, and, and so forth. I, I wanted to bring it back to, to one of the slides that, that Jorge had about, um, uh, about gender and discrimination and so forth and say that I think that, um, There's one issue is what you can do to sort of improve entrepreneurial ability among people who are currently entrepreneurs, but I also think we shouldn't lose sight of the fact that we, we, and a lot of these economies, we're not doing a very good job of selecting who becomes an entrepreneur to begin with. And we need to think hard about that, about that, uh, that, uh, question as, as, as, as well. And, and, um, You know, we get to better entrepreneurs, uh, either from, from better selection or from, or from, or from better training. And then, on administrative data, so, so funny, I've spent most of my career kind of justifying why survey data should be thought of as as reliable as administrative data. So it's, so I, I, you know, I'm, I'm happy to hear, uh, people say that, well, maybe, and, and it, and it's clearly gonna be the case when we think about taxes and, and, and, you know, reasons that people would, would, would skew administrative data. I, I think I think that the data are valuable, um, and I think the research that's been, that's been done when thinking about firm to firm networks and how networks, uh, that, that, that, the data that, that, um, that, that, that the VAT data in particular allow us to, um, the, the patterns of people that, that those data allow us to understand is, is, is, uh, I'm gonna say unique cause it's really difficult. It would be really difficult to, to, to, uh, uh, generate any kind of broader, um, Uh, survey. Um, and, uh, in, in a broader sense, in a, in, in a survey. And, um, and then the question of, of representation of the countries where these data are available. If I go back 10 years, I'm, I don't think VAT data were available in any country. Now, we've got, you know, 4 or 5 countries in Africa where, um, uh, you know, where there are at least 3 countries where they're very broadly available and, and, and, uh, And Ethiopia as well, where people are using them. So, I, I think that it's likely that perhaps by demonstrating that other countries will, will be, will begin to, will, will begin to see these, these, these, these kinds of data pop up in other places. So, um, I, I, I, I Take these, I take these points, but I, but I think that um there are a set of questions that I think it's gonna be quite challenging for us to, to address with, uh, uh, with, uh, with, with survey data where I see uh promise in those, uh, in, in those, in those data. And thanks. Thanks, Chris. Deborah. On my, on my side, maybe, um, one of the things that I was uh trying to pass as a message is also the fact that that combining data, data, uh, it's, uh, extremely valuable. On our side, what we do is we combine the balance sheet information of the firms that are. Uh, externally, so we have hard data combined to survey data and then matching with, uh, all the data, data sources, databases on the location of the firms, and, uh, all information that you have on the overall external environment and I think it's uh the combination of information. So that brings a lot of value in understanding because then you can understand how much is the firm behavior, how much is the external environment, how much is the policy intervention that influence what the sing, what motivates the action of the firm. So I think that particularly in topics extremely difficult to analyze like the the the green transition where data, data are difficult, the combination of different ways of seeing the same phenomenon, including the survey data is important. Thanks, Deborah. Uh, so, Dean, over to you for the last word. Sorry, sorry. Um, uh, yeah, no, I think this is, you know, for what it's worth, I think this discussion on data has been fascinating, and I, I, um, haven't worked as much with it personally, with the kind of administrative data, but I think it, it hits on some of the, I, I, I'm, I'm excited by the, by the promise of the integration, both because of the scale, administrative data possibilities and And also because I think of the importance in trying to get at some of the, the general equilibrium issues that I, that I was talking about, um, that, that is, you know, clearly a, a path to trying to, um, to try to tackle some of those. So it was, uh, thanks for having me on this and I was, you know, really was, it was great listening to everybody and learning more about what everybody's um up to and where everybody is thinking there's big gaps and rooms for progress. So I'm excited for the next 20 years. Thanks very much, Dean. That's actually sort of a segue into what I wanted to say by um to, to wrap up this. So this is not just wrapping up this session in today's event, but also the series of 3 events that we've put on to, to celebrate these 20 years of KCP. I keep focusing on the 20 years, but I occasionally drop in and $80 million of over $80 million of donor funded research uh during this time. So, this has really been a very significant effort on the part of our KCP donors, which I think over the years we've had over 20 different donors to KCP who have contributed, uh, to make a lot of this research possible and to advance on, on, on this agenda. I did want to particularly acknowledge that our current donors to the, the current round of KCP, so, uh, CETA and Sweden. Uh, um, France, uh, the government of Japan, and also the EU through two different, uh, uh, DGs that have been contributing to, uh, to KCB funded research. Uh, um, I think this is, uh, it's a great foundation for the next 20 years and we're looking forward to being able to do much more. Um, just to close up, I also did want to give a big thank you to Karina and Bintao for all the work that they've done in, uh, in putting together again, not just today's event, but the 3 events and for making KCP run. Also, this is now becoming more of an inside the bank comment for a second, but I did also want to, uh, thank Bintao in particular, who's been with KCP for as long as I can remember. Although my memory is failing these days, so, you know, maybe that isn't such a big compliment, but Bintao is going to be moving on shortly to, uh, do a new opportunity in DFI, right? No. Water, sorry, in the water global practice and sitting beside Bintao and just to introduce quickly and we'll do it more systematically later, but we have Anna Bokina who is joining us from the Africa region, who will be stepping into Bintao's role. So, um, welcome, uh, welcome Anna as well. So with that, um, uh, uh, let me again thank our speakers, David, Jorge, and our, our panelists, Deborah, Dean, and, uh, and Chris, and, uh, you know, big round of applause and everybody for all of their, uh, their hard work to make this happen and for all the interesting insights. Thank you very much.
showAllTranscripts
no
duration
PT1H16M11S
scene7File
worldbank/KCP_III_private_sector
scene7Domain
https://worldbank.scene7.com/
scene7FileAvs
worldbank/KCP_III_private_sector-AVS
title
KCP III private sector
description
KCP III private sector
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
lp-heading-top-medium
lp-heading-bottom-medium
col-xs-12
col-sm-12
col-md-2
col-lg-2