Pensions and Informality

videoType

dynamic-media

videoDmUrl

keyFrameImage

timestamp

00:00 Um,

00:01 Welcome.

00:02 Uh,

00:02 my name is Dion Film.

00:03 I'm the director of the World Bank's research department.

00:06 So,

00:06 uh,

00:07 welcome to this October edition of our policy research talks.

00:10 Um,

00:11 as some of you may know,

00:12 these,

00:12 uh,

00:12 provide us an opportunity to present work coming out

00:14 of the department with the goal of sharing the findings

00:17 inside and outside the department as well as with others outside of the World Bank.

00:21 Uh,

00:22 with that,

00:22 I'd like to welcome our audience in the room today,

00:25 uh,

00:25 on the Webex as well as on YouTube.

00:28 Uh,

00:29 today my colleague,

00:30 uh,

00:30 Clement Joubert is going to,

00:31 who is an economist in the human development team in the research department,

00:35 he's going to discuss,

00:37 uh,

00:37 how his and other

00:39 research informs pension policy design

00:42 in high informality settings,

00:44 and he's gonna cover issues such as the disincentives for formal work,

00:48 gender disparities,

00:50 and the extent to which workers in the

00:51 informal economy can participate in voluntary pension schemes.

00:56 Uh,

00:56 it wouldn't be a research talk without a

00:58 conclusion that points to further research needs,

01:00 so I think he's going to conclude

01:01 with some unanswered questions and knowledge gaps.

01:05 Um,

01:05 it's my great pleasure to welcome,

01:07 uh,

01:07 Mikhael Ruovsky as a discussion today,

01:10 not least because he was one of the very first people

01:13 I worked with when I joined the World Bank to work on the 1995

01:17 edition of the World Development Report entitled Workers in an Integrating World.

01:22 Um,

01:22 Michel is the global director for Social Protection and Jobs,

01:25 overseeing the World Bank's work in developing

01:27 systems that protect the poorest and vulnerable

01:30 from crises and shocks,

01:32 as well as supporting private sector-led growth.

01:35 Until July 2016,

01:37 he was the director for multilateral organizations,

01:40 and prior to that he was the country director for the Russian

01:42 Federation and the World Bank's resident

01:44 representative in Moscow for three years.

01:47 Before that,

01:48 he held several positions including sector manager for social protection

01:51 as well as country economist for the Russian Federation,

01:54 in addition to the

01:55 pinnacle of his career,

01:56 which was being a core team member of the 1995 WDR.

02:00 Uh,

02:00 prior to,

02:01 uh,

02:01 working,

02:02 uh,

02:02 to joining the World Bank,

02:03 and this is actually very relevant to today's,

02:06 uh,

02:06 discussion,

02:06 he helped design the Polish pension system while serving as the

02:09 director of the Office for Social Security Reform in that country.

02:13 He also taught at the Institute of Economic

02:15 Policy at the Warsaw School of Economics.

02:18 So the session today,

02:19 I'll ask

02:21 Clement to talk for approximately 45 minutes,

02:23 after which we'll hear from Michel for about 10 to 15 minutes,

02:26 uh,

02:26 to get his reflections.

02:28 We'll conclude the session with Q&A from the audience.

02:31 Um,

02:31 if you have a question,

02:33 please use the raised hand option in Webex

02:35 or signal to me in the chat that you have a question,

02:37 and I can call on you.

02:39 If you're in the room,

02:39 just raise your hand.

02:41 Um,

02:42 if you're on YouTube,

02:43 please put a question in the YouTube chat.

02:46 I won't see it,

02:46 but it will get relayed to me.

02:48 So,

02:48 uh,

02:48 I can,

02:49 I can,

02:49 uh,

02:50 relay that to,

02:51 to Clement

02:52 or Michel.

02:53 Um,

02:54 just a reminder,

02:55 we are recording the session,

02:56 um.

02:58 And also,

02:58 if you're online,

02:59 please mute if you're not speaking.

03:01 So with that,

03:02 let me turn it over to to Clement.

03:12 All right,

03:12 um,

03:13 is the microphone working

03:14 properly?

03:15 No,

03:17 hello.

03:17 Oh,

03:17 there you go.

03:19 Um,

03:19 thank you for the introduction,

03:21 Dion.

03:22 Uh,

03:22 thank you,

03:23 Michel,

03:23 for agreeing to discuss this,

03:25 this talk,

03:26 and,

03:26 uh,

03:27 good morning,

03:28 everybody here in the room and,

03:29 um,

03:30 and online.

03:31 Um,

03:33 so my name is Clement Joubert.

03:34 I'm in the,

03:35 in the research group,

03:36 um,

03:37 and today I'm gonna draw from three

03:40 of my research papers.

03:42 Uh,

03:42 to discuss

03:43 issues related to pension design in context where informality is prevalent.

03:51 Um,

03:52 You probably are aware that that

03:57 most workers globally do not contribute to pensions.

04:01 In fact,

04:02 about 2/3 of workers worldwide

04:04 do not contribute towards an old age pension.

04:07 Um,

04:07 and at the same time,

04:08 uh,

04:09 we know that low and middle income countries are aging,

04:12 and they're aging faster than,

04:15 uh,

04:15 current developed countries did in the past.

04:18 So we can expect in the near future that fewer active individuals

04:23 will have to support a growing elderly population.

04:26 Um,

04:27 and that this will put a strain

04:30 on traditional old age,

04:32 uh,

04:32 support arrangements.

04:34 So,

04:35 um,

04:35 encouraging long term savings,

04:37 uh,

04:38 for example,

04:39 through,

04:39 through

04:40 by extending pension coverage,

04:41 uh,

04:42 could be a way to,

04:43 uh,

04:43 finance the future costs that are associated with an aging population.

04:49 Um,

04:50 now,

04:50 the picture is actually grimmer than those numbers,

04:53 um,

04:53 um,

04:55 suggest because,

04:56 uh,

04:56 when you start looking at,

04:58 um,

04:59 at how people contribute over their life cycle,

05:01 you find that there's even fewer,

05:03 uh,

05:04 consistent contributors.

05:06 Uh,

05:06 pension economists,

05:08 uh,

05:08 talk about contribution densities,

05:10 uh,

05:10 to describe

05:11 the fraction of the months since,

05:14 uh,

05:14 the,

05:14 uh,

05:15 since a worker enters the labor force

05:17 where,

05:17 uh,

05:18 they contributed.

05:20 And um if we look at

05:22 data from uh from Chile,

05:24 so on the right you see uh

05:25 the distribution of these contribution densities among men

05:29 and women um in Chile,

05:31 you can see,

05:32 um,

05:33 you can see bars at 0 and 1.

05:35 Those are people who never contribute or who always contribute,

05:39 but you also see a lot of people who,

05:41 um,

05:41 who switch in and out of contributing.

05:45 Um,

05:45 so a lot of switchers for women,

05:48 you know,

05:48 labor market exists,

05:49 exits further reduced contribution densities,

05:52 um,

05:53 and so that means that if in the cross sections,

05:55 or,

05:55 so,

05:56 for example,

05:56 in Chile you have about 2/3 of workers who contribute in a given year,

06:00 uh,

06:01 once you start looking at people who contribute,

06:03 uh,

06:03 year,

06:03 year in,

06:04 year out,

06:04 you're left with maybe 1/3 of,

06:06 of,

06:06 um,

06:07 of,

06:07 uh,

06:08 individuals contributing.

06:12 Um,

06:13 another difficulty with pension designs in,

06:15 in this kind of,

06:16 uh,

06:16 setting is that you have what you could call a vicious circle

06:20 of contribution evasion.

06:22 Um,

06:23 so obviously pensions,

06:24 um,

06:24 ask you to pay a cost today in exchange for a benefit far in the future.

06:29 And so as soon as you have,

06:31 uh,

06:31 any,

06:32 any form of present bias or even,

06:34 um,

06:34 just liquidity constraints,

06:36 uh,

06:36 this is going to create incentives for workers to avoid those,

06:40 uh,

06:40 those payments.

06:42 And once you start having contribution evasions,

06:45 you're going to get a pension system that produces

06:47 low pensions or maybe that is in deficit.

06:52 And then what you need to do in that case

06:54 is either you're going to have to increase the,

06:56 the contribution rate to prop up pensions or you're

06:58 going to have to raise money through through taxes

07:01 and in both cases that can further

07:03 uh uh um aggravate

07:06 the evasion and that was very much the situation

07:09 uh in Chile before the famous privatized pension system

07:13 was,

07:14 was implemented in 1981.

07:17 Um,

07:19 So

07:20 in this talk I'm going to consider

07:23 3

07:24 main questions,

07:25 um,

07:27 drawing from from 33 research papers.

07:30 So first I'm going to present some results

07:31 on the extent to which pension pension system rules

07:35 generate informality themselves.

07:38 Um,

07:39 second,

07:39 I'm going to talk about,

07:41 um,

07:41 gender pension gaps and how the design of the pension system can,

07:47 uh,

07:47 can help reduce them.

07:49 And lastly,

07:50 I will talk about

07:52 um

07:53 whether informal workers

07:55 in the absence of pensions are

07:57 uh saving uh accumulating wealth or saving um or not.

08:03 Um,

08:04 so the first two,

08:05 questions will,

08:06 will be in the context of Chile,

08:08 um,

08:08 which is at the lower end of the spectrum in terms of informality rates,

08:12 but,

08:13 as we've seen they grapple with very,

08:15 very much with these issues of pension coverage,

08:17 um,

08:18 because of these,

08:18 uh,

08:19 inconsistent,

08:20 uh,

08:20 contribution patterns,

08:22 and the third,

08:23 uh,

08:23 paper will draw,

08:24 uh,

08:24 uh,

08:25 we use data from Pakistan which is,

08:27 uh,

08:27 uh,

08:28 uh,

08:28 a setting where informality is nearly universal.

08:33 Um,

08:34 and,

08:34 um,

08:35 and,

08:35 um,

08:36 you know,

08:36 so the,

08:36 the second paper is,

08:37 is,

08:38 uh,

08:38 co-authored with Petra Todd at the University of Pennsylvania,

08:40 and the third paper is,

08:41 is,

08:42 uh,

08:42 co-authored with,

08:43 with,

08:43 uh,

08:44 Priyanka,

08:44 uh,

08:44 who's here in the room.

08:45 So thanks for coming.

08:48 Um,

08:49 before I get into the results,

08:50 I would like to,

08:51 to,

08:51 uh,

08:52 introduce,

08:52 um,

08:53 a,

08:53 a small framework to,

08:55 to,

08:55 to show how,

08:56 how

08:57 I think about,

08:58 um,

08:58 the choice between,

08:59 uh,

09:00 formal,

09:00 formal and,

09:01 uh,

09:01 and informal jobs,

09:03 uh,

09:03 and that will help us,

09:04 uh,

09:05 understand the,

09:05 the results later.

09:07 Um,

09:08 so if we consider a worker who is either choosing,

09:12 choosing between a formal job and an informal job opportunity,

09:16 or maybe,

09:17 um,

09:17 you know,

09:17 he's trying to decide whether to register his business or not,

09:21 um,

09:22 we can

09:23 write the value of working in,

09:25 um,

09:26 let's say in the formal sector,

09:27 um,

09:28 as,

09:29 um,

09:30 including three main components.

09:32 Um,

09:33 so the first one is the utility derived by,

09:35 by the wages that you can,

09:37 uh,

09:37 obtain in either,

09:39 uh,

09:39 through other,

09:40 either option,

09:41 but there may be also a lot of non-wage considerations.

09:45 Uh,

09:45 you might

09:46 think that some people argue that the informal sector gives you more flexibility.

09:51 Uh,

09:51 on the other hand,

09:52 formal jobs might give you access to amenities or,

09:54 or,

09:55 uh,

09:55 or to,

09:56 to,

09:56 um,

09:57 let's say health insurance,

09:58 so you might have these non-wage preferences for working

10:01 for,

10:01 uh,

10:01 one sector versus the other,

10:03 and these might different,

10:04 uh,

10:04 for different workers.

10:06 And lastly,

10:06 you're going to have dynamic considerations.

10:09 So if you work formally today,

10:10 that's going to have consequences for you in the future as well.

10:14 Um,

10:14 you might accumulate different sets of skills.

10:17 Um,

10:18 you're going to accumulate pension savings or pension rights,

10:21 uh,

10:21 if working formally.

10:23 Um,

10:24 it might be the case that if you work formally today,

10:26 that's going to make it easier for you to keep working formally tomorrow,

10:29 so that's going to increase your,

10:30 your,

10:30 your options in the future.

10:32 And um you might,

10:33 um,

10:34 it might help you um achieve eligibility for some um government transfers.

10:40 Now,

10:41 if we zoom in on the the wage part of um of of this equation,

10:46 um,

10:47 so obviously if you work formally,

10:49 your disposable wage is going to be reduced in proportion to the contribution,

10:52 the pension contribution rate because that money is going to be

10:55 uh either stored away in an individual account or is going

10:57 to be collected by the government in exchange for future pensions.

11:01 Um,

11:01 and then,

11:02 um,

11:03 your earning opportunities in formal or,

11:05 or informal jobs will depend on the type of skills that you have,

11:08 both schooling,

11:09 um,

11:10 the experience you've accumulated in either sector,

11:12 um,

11:13 also you might have an innate ability

11:15 to work,

11:16 um,

11:16 to be productive in,

11:17 in,

11:17 in,

11:18 um,

11:18 in formal or informal jobs,

11:20 and then you also have,

11:22 you know,

11:22 um,

11:23 time varying opportunities that,

11:24 that arise,

11:25 uh,

11:25 in,

11:25 in either sector.

11:26 So all of this to say that the value of working in either sector will be

11:31 heterogeneous in the population.

11:34 Um,

11:34 and so in when we think about the impact of pension rules on these decisions,

11:39 what is going to matter is,

11:41 is the relative value,

11:43 evaluation of formal versus,

11:45 uh,

11:45 informal wage opportunities for each worker,

11:47 um,

11:48 in each point in time.

11:50 And,

11:50 um,

11:51 so this is,

11:52 uh,

11:52 written as VF minus VI.

11:55 Um,

11:56 and you might have different,

11:57 uh,

11:58 you could imagine different cases,

11:59 right?

11:59 And on,

12:00 on the left hand side,

12:01 you have a situation where workers are relatively homogeneous

12:04 and they are,

12:05 um,

12:06 close,

12:07 relatively close to the indifference point

12:09 between working formally or informally.

12:11 Um,

12:12 whereas in case two you have workers who are heterogeneous in how they value,

12:18 uh,

12:18 each sector,

12:19 uh,

12:19 with some workers who are,

12:21 who value formality very much,

12:23 maybe because they can earn a lot more in the

12:26 formal sector or because they value health insurance a lot,

12:29 um,

12:29 and other workers who,

12:31 um,

12:31 value informality a lot,

12:33 and that could be,

12:33 for example,

12:34 because they don't even have access to formal jobs,

12:36 right?

12:36 That could be a way to describe it.

12:39 If we introduce,

12:40 if we change the rules of,

12:41 of the pension system in a way that makes formality a little bit more attractive,

12:46 um,

12:47 in case one,

12:48 you're likely to displace

12:49 a large number of workers

12:51 into formality,

12:53 whereas in case two,

12:55 the margin is going to be much thinner and,

12:56 and,

12:57 um,

12:57 you can expect your,

12:58 your effects to be,

12:59 uh,

13:00 much lower.

13:00 So

13:01 it all depends on how these,

13:02 these,

13:02 um,

13:03 um,

13:04 productivities and,

13:05 and preferences are distributed in the population.

13:09 Um,

13:10 and last little bit before I move into,

13:12 uh,

13:12 move on to results,

13:13 um,

13:14 the results,

13:15 um,

13:15 for the first two papers at least,

13:17 um,

13:17 are,

13:18 um,

13:19 are,

13:19 uh,

13:20 use a methodology called,

13:21 um,

13:21 a structural estimation,

13:23 and

13:24 without going into details,

13:25 I just want to give you a sense of,

13:26 of,

13:26 of what it is so that you,

13:28 you can interpret the results better.

13:30 Um,

13:31 what,

13:31 uh,

13:32 you do first is you develop a model.

13:33 In this case it's a model of how,

13:35 uh,

13:36 individuals or households are saving

13:38 and choosing and,

13:39 and,

13:40 and supplying labor over their life cycle.

13:42 And the key,

13:43 as we've just discussed,

13:44 is to introduce enough heterogeneity into the model that you can capture

13:49 this distribution

13:50 of,

13:50 of how people value formality versus informality.

13:55 Um,

13:55 then you estimate the parameters of the model using,

13:58 um,

13:59 panel data in this case of on Chilean households,

14:03 and once you have this,

14:04 this,

14:04 this,

14:05 this,

14:05 uh,

14:06 this tool,

14:06 this estimated model,

14:07 you can,

14:08 um,

14:09 because the

14:10 rules of the pension system are explicitly modeled,

14:13 you know,

14:13 the,

14:14 the pen,

14:14 the pension system contribution rate is a parameter in the model,

14:17 you can switch those pension rules or move these pension rules around

14:21 and simulate what happens to labor supply saving decisions.

14:24 Uh,

14:24 but also pension benefit levels and government costs under different design.

14:31 OK,

14:31 so,

14:32 um,

14:33 1st,

14:34 1st

14:35 question that,

14:36 uh,

14:36 we're going to look like,

14:37 look at is,

14:39 um,

14:40 how do pension rolls,

14:41 uh,

14:42 do pension rolls rather,

14:43 um,

14:44 contribute to,

14:45 to informality?

14:47 Um,

14:49 So what we do here is,

14:51 um,

14:54 using this kind of estimated structural model,

14:56 we're going to change the contribute,

14:58 we're first going to change the contribution rate around.

15:01 Um,

15:01 and the reason we're interested in the contribution rate is because,

15:05 um,

15:05 you know,

15:06 contributions,

15:06 pension contributions are essentially

15:08 mandatory

15:09 and illiquid savings,

15:11 and to the extent that,

15:12 um,

15:13 you're forcing people to save more than they would have,

15:15 um,

15:16 on their own,

15:17 this is going to,

15:18 um,

15:19 create this utility and essentially reduce

15:21 the value of working formally.

15:24 Um,

15:25 so

15:25 on the right hand side you can see the,

15:28 the effect of moving the contribution rate around

15:31 from,

15:31 uh,

15:31 a value of 5%,

15:32 which it would be a very low contribution rate,

15:35 um,

15:36 all the way to 20%,

15:37 um,

15:38 so that kind of spans,

15:39 uh,

15:40 the,

15:40 the typical contribution rates you see around the world and

15:43 the,

15:43 um,

15:44 Chilean,

15:44 in the Chilean case,

15:45 which is the,

15:46 the source of these,

15:47 these numbers,

15:48 um,

15:48 the,

15:49 the contribution rate is at 10%.

15:51 And what you can see is that in the left hand side of the graph,

15:54 the first margin that is activated as you move the contribution rate around

15:59 is the substitution between pensions and non-pension saving.

16:03 So the first thing that that uh households can do is,

16:06 uh,

16:06 if you,

16:07 if you ask them to contribute more

16:08 towards pension they might reduce their savings elsewhere

16:12 but uh at some point,

16:13 uh,

16:14 they're going to either run out of savings to substitute from

16:17 or they are going to want to keep

16:19 a liquid savings so that they can.

16:21 Um,

16:22 self-insure against,

16:23 against short term shocks

16:24 and so that margin kind of shuts down and what you start seeing is

16:28 more people switching to,

16:30 um,

16:30 to informality

16:32 and,

16:32 and,

16:33 or households just producing their current consumption.

16:37 So overall,

16:38 uh,

16:38 the impact that we find of increasing,

16:41 uh,

16:41 the contribution rate by 10% points to give you an order of magnitude is,

16:45 uh,

16:46 an increase in informality by about 8.3% points for men

16:50 and 5.3% points for women,

16:52 um,

16:53 so.

16:54 You know,

16:56 essentially what that means is that this is,

16:57 these are significant or sizable numbers,

16:59 but at the same time

17:00 it also says that,

17:02 you know,

17:02 the informality we see in Chile is not all

17:04 generated,

17:04 they're generated by the,

17:06 by,

17:06 by,

17:06 by the,

17:07 the,

17:07 the obligation to contribute.

17:10 Um,

17:11 one interesting bit as well is that these effects are concentrated,

17:15 uh,

17:15 among workers with lower contribution density.

17:18 So,

17:19 uh,

17:19 essentially

17:20 workers who always contribute,

17:22 uh,

17:22 that the fraction of workers who always contribute does not change,

17:25 um,

17:26 and really it's those occasional contributors who are,

17:29 are pushed out of,

17:30 uh,

17:30 formality.

17:34 Um,

17:35 a second

17:36 feature of the pension system that has been,

17:38 um,

17:39 accused or,

17:40 or,

17:40 or,

17:40 or,

17:40 um,

17:41 thought to,

17:42 um,

17:43 generate informality is minimum pensions.

17:46 Um,

17:47 and so the mechanism here is that minimum pensions implicitly

17:50 tax contributions.

17:52 So this is,

17:53 uh,

17:53 easy to see

17:54 when you,

17:55 when you consider a very simple minimum pension.

17:57 Um,

17:57 so,

17:58 you know,

17:58 the,

17:59 the black line on this graph is

18:01 your contributory pension.

18:02 So the more you contribute,

18:03 the higher your,

18:04 your pension,

18:05 and let's say the government wants to guarantee that pensions

18:08 are not going to drop below a certain minimum.

18:11 Um,

18:12 in that case,

18:13 um,

18:14 for people who expect to qualify for that minimum pension,

18:17 um,

18:17 it's,

18:18 it's quite obvious that

18:19 contributing one extra year is not gonna raise their pension level,

18:23 and in that sense,

18:25 those contributions,

18:26 those pension contributions are taxed,

18:28 uh,

18:28 implicitly at a 100%,

18:31 uh,

18:31 rate.

18:32 Um,

18:33 so that's going to create disincentives for,

18:35 uh,

18:36 this segment of workers to,

18:37 um,

18:38 to work formally.

18:41 Um,

18:41 and so it's been proposed and in fact implemented in,

18:44 in,

18:44 in some reforms,

18:45 uh,

18:46 to,

18:46 to try to reduce these implicit tax rates,

18:49 um,

18:50 and so the way,

18:51 uh,

18:51 you can do that

18:52 is by,

18:53 um,

18:54 tapering the minimum pension benefit and what we mean by that is,

18:58 is moving from this,

18:59 uh,

18:59 red dashed

19:01 line to the solid line.

19:03 So what,

19:03 uh,

19:04 what you're doing here is,

19:05 um.

19:06 As

19:07 people contribute more,

19:08 you're reducing the government transfers,

19:10 so you're giving them a little bit less,

19:12 but it's not a one for one substitution.

19:14 So you,

19:15 as they contribute more,

19:16 they're still increasing their,

19:17 their pension a little bit,

19:19 but they're still implicitly taxed,

19:21 uh,

19:21 at a certain rate.

19:24 Um,

19:25 um,

19:26 and so,

19:27 uh,

19:27 we're going to anchor our analysis again on,

19:30 on the Chilean case where in 2008,

19:33 um,

19:34 the minimum pension was reformed and they adopted,

19:38 um,

19:38 a design where

19:40 the implicit tax rate is about 30%.

19:45 So,

19:45 and

19:46 uh the policy experiment that we do specifically is that we're going to fix

19:50 um the level of um

19:53 the intercept of that of that graph so the level of the basic pension

19:56 and we're going to move

19:57 the slope around so vary the implicit tax rate and see what happens to to informality

20:03 and what we find is that actually

20:04 um the impact is very small

20:07 so informality doesn't change that much.

20:10 So this,

20:11 which is kind of contrary to the,

20:13 the uh ex ante expectation.

20:15 So what is happening here,

20:16 um,

20:17 it's true that you are lowering disincentives for people who would have,

20:21 um,

20:21 um,

20:22 qualified for the old minimum pension.

20:25 Um,

20:26 but at the same time,

20:27 now you're going to have more recipients,

20:28 right?

20:28 It's pretty obvious from the,

20:30 from the graph as you move from the dashed,

20:32 uh,

20:32 red line to the solid line,

20:34 you're

20:34 increasing the range of,

20:36 of,

20:37 uh,

20:37 of,

20:37 uh,

20:39 of,

20:39 or the,

20:39 the,

20:40 the,

20:40 the set of people.

20:41 Who are going to qualify for a minimum pension.

20:43 So essentially what you're doing is you're

20:46 spreading this implicit tax rate over more people.

20:49 So you're reducing the tax rate,

20:50 but you're,

20:51 you're,

20:51 uh,

20:51 incorporating more and more people into these incentives.

20:53 And so even theoretically the effect is,

20:56 um,

20:57 is,

20:57 uh,

20:58 is not,

20:58 uh,

20:58 is,

20:59 is ambiguous.

21:00 On top of that,

21:01 you might expect that um

21:04 contrary to what's happening with the contribution

21:06 rate that we talked about before,

21:08 this is um

21:09 this is very far in the future for many workers,

21:11 right?

21:11 So if you're 20 years old and you're trying to decide between formal and informal,

21:15 you,

21:15 there's still a lot of uncertainty uh about your career that uh prevents you from

21:19 knowing exactly if you are going to qualify for a minimum pension or not,

21:23 uh,

21:23 even,

21:23 even in the absence of any,

21:25 um,

21:26 present biased or anything like that.

21:28 Uh,

21:28 so this uncertainty is going to dilute

21:30 these disincentive,

21:31 uh,

21:31 especially at younger ages,

21:33 and the last little bit that explains this result is that,

21:36 um,

21:36 you are transferring more money,

21:38 uh,

21:39 to the,

21:40 to,

21:40 to from,

21:41 from the government,

21:42 as you can see,

21:42 you know,

21:43 the,

21:43 the,

21:43 the government transfers can be loosely captured by the shaded area

21:47 below the minimum pension line and above the contributory pension,

21:52 so you can see that as you move,

21:54 as you,

21:54 as you,

21:55 um.

21:55 You make the,

21:56 the,

21:57 you increase the slope of the minimum pension,

21:59 you're going to transfer more money to the population that's going to create

22:02 wealth effects and disincentivize um uh formal work.

22:07 All right,

22:08 so just to summarize this,

22:09 uh,

22:10 the,

22:10 the,

22:10 the set of results,

22:12 um,

22:12 um,

22:13 if we look at the contribution rate,

22:14 we do see,

22:15 uh,

22:16 we do predict a sizable effect,

22:18 uh,

22:19 on informality,

22:20 an effect that is

22:21 likely larger for younger workers who are not yet

22:23 locked in either the formal or the informal sector.

22:27 Uh,

22:28 note that obviously informality is not the only and perhaps not even

22:31 the main consideration in deciding what your contribution rate should be.

22:35 You know,

22:35 pension adequacy would be the first,

22:37 uh,

22:37 thing that comes to mind,

22:38 but also maybe more subtly,

22:40 um,

22:40 if you lower the contribution rate,

22:42 maybe you're going to reduce informality a little bit,

22:45 but you also are going to increase the cost of your safety net because

22:49 people are going to accumulate less money or less pension rights into the system,

22:52 and so they're more likely to qualify for whatever minimum pension you have there.

22:57 Um,

22:58 on the other hand,

22:59 uh,

22:59 minimum pensions

23:01 and particularly the,

23:02 the,

23:03 the implicit tax rate or the way you're designing them,

23:06 um,

23:06 is likely to have a small effect on,

23:08 on informality,

23:10 um,

23:10 and if to the extent that they exist,

23:12 they are going to be close to,

23:14 to the retirement age when people know exactly,

23:16 um,

23:17 or know much more about their,

23:19 their eligibility.

23:21 Um,

23:22 other considerations,

23:23 as I said,

23:23 you're transferring a lot more money,

23:25 uh,

23:25 to,

23:26 to people under a more tapered design,

23:28 and so that's going to have a first order impact on the cost of your,

23:32 your minimum pension

23:33 for a given,

23:34 uh,

23:34 you know,

23:34 consumption floor that you're trying to guarantee.

23:39 All right,

23:40 so 2nd,

23:40 uh,

23:41 2nd set of results,

23:42 now I want to talk about,

23:44 uh,

23:44 gender pension gaps.

23:46 And this is drawing from a,

23:47 from a different paper but still in the context of,

23:50 uh,

23:50 of Chile.

23:52 Um,

23:53 so

23:53 gender

23:54 pension gaps are pervasive.

23:56 They,

23:56 they are,

23:57 they exist in most,

23:58 uh,

23:58 pension systems

23:59 around the world,

24:00 and

24:01 the reason is because there are,

24:02 um,

24:03 several compounding factors that,

24:04 that contribute to them.

24:06 Um,

24:07 the first one is labor market opportunities.

24:10 Women tend to be paid,

24:11 uh,

24:12 lower wages.

24:13 They also,

24:14 um,

24:15 may have,

24:16 uh,

24:16 lower access to formal job opportunities.

24:20 Um,

24:21 also,

24:21 um,

24:22 women tend to exit the labor force when they have children,

24:26 and that's going to reduce their,

24:27 um,

24:28 their contribution densities.

24:31 But also,

24:32 um,

24:32 uh,

24:33 there are rules in pension systems that

24:35 are,

24:35 uh,

24:36 not gender neutral.

24:37 Um,

24:38 so many,

24:38 not all,

24:39 but many,

24:40 uh,

24:40 systems have different retirement ages for women versus men.

24:45 Um,

24:45 and particularly in DC schemes,

24:48 um,

24:49 uh,

24:49 the,

24:50 the,

24:50 the,

24:51 the pensions might be calculating using gender specific

24:54 mortality tables,

24:56 and because women,

24:57 uh,

24:57 live longer,

24:58 that will result in lower,

24:59 uh,

24:59 pension benefits for them.

25:03 Um,

25:03 so we're going to examine two strategies for tackling the gender,

25:08 uh,

25:08 pension gap.

25:10 And the first one is

25:12 to just expand the,

25:13 the minimum pension,

25:14 um,

25:15 because women are more likely to have few years of

25:17 contribution and low pensions they are going to to benefit

25:21 a lot more from this type of,

25:22 of policies

25:23 and in fact that's what happened in uh in Chile in 2008.

25:26 The,

25:27 the system was reformed.

25:28 Um,

25:29 targeting,

25:30 uh,

25:30 specifically gender pension gaps,

25:32 um,

25:33 and,

25:34 um,

25:34 the way they,

25:35 they,

25:35 they,

25:36 um,

25:36 they try to achieve it is by increasing the minimum pension level,

25:40 removing contribution requirements,

25:42 and,

25:43 um,

25:43 also widening the means test

25:45 that,

25:46 uh,

25:46 um,

25:47 that you need to pass to,

25:48 to be eligible.

25:50 Um,

25:51 and,

25:52 uh,

25:52 this was,

25:53 uh,

25:53 this,

25:53 um,

25:54 uh,

25:55 effectively reduces the gender pension gap,

25:57 um,

25:58 by,

25:58 by,

25:58 um,

25:59 by a large amount,

26:00 almost,

26:00 uh,

26:01 um,

26:02 uh,

26:02 splits,

26:03 splits it in half.

26:04 So,

26:04 um,

26:05 going from 44%,

26:07 uh,

26:07 gap to 23% gap.

26:09 And we also find that the effect is largest for women

26:14 with lower wages and lower attachment to

26:18 to formal work.

26:19 So,

26:20 um,

26:21 women with lower schooling,

26:22 uh,

26:23 single women

26:24 will um

26:25 will see the biggest reduction in the gender pay gap.

26:29 Um,

26:30 the second

26:31 strategy that we,

26:32 uh,

26:32 simulate,

26:33 and that one was not,

26:34 didn't make it into that reform,

26:36 is to,

26:37 um,

26:38 to equate retirement ages.

26:41 Um,

26:42 so,

26:43 uh,

26:43 currently women in Chile can claim a,

26:45 a pension

26:46 at 60,

26:47 um,

26:48 instead of 65 for men,

26:50 and we look at what happens if we equate those two,

26:52 those two numbers,

26:54 and we find that the reduction in the gender pay gap is also very large,

26:57 uh,

26:58 in fact,

26:58 of the same order of magnitude than what,

27:00 uh,

27:01 we found,

27:02 uh,

27:02 um,

27:03 with the minimum pension,

27:04 uh,

27:05 um,

27:06 policy experiment.

27:08 Uh,

27:08 but now,

27:08 uh,

27:09 it's a very different,

27:10 uh,

27:10 groups of women are affected.

27:12 So now,

27:12 uh,

27:13 you're reducing the gap among women with higher schooling,

27:16 um,

27:17 and,

27:17 and,

27:18 and,

27:18 uh,

27:18 and single women.

27:20 Um,

27:21 and the mechanisms that explain this impact is that

27:24 when you,

27:25 uh,

27:25 uh,

27:26 increase the retirement age,

27:28 first you're reducing longevity at retirement.

27:30 So,

27:31 um,

27:31 in,

27:32 in systems like Chile's,

27:33 you,

27:34 you have a,

27:34 you accumulate a certain certain amount of money and then you buy an annuity

27:38 or something,

27:39 uh,

27:40 similar to that.

27:41 So the longer you're expected to live,

27:43 the,

27:44 the,

27:44 the smaller your,

27:45 your,

27:45 um,

27:47 your,

27:47 your benefit.

27:48 Uh,

27:49 in addition,

27:49 you might expect that between 60 and 65,

27:51 some,

27:52 some women are gonna work and work formally,

27:54 so they're going to contribute to their,

27:55 to their account,

27:56 but that's actually not the main,

27:58 uh,

27:58 the main,

27:58 uh,

27:59 explanation.

28:00 And then lastly,

28:01 uh,

28:02 pension savings are left for 5 more years,

28:04 uh,

28:05 to accrue interest,

28:05 and that's also going to increase,

28:07 uh,

28:07 increase pensions,

28:09 um.

28:12 Right,

28:12 so,

28:13 um,

28:14 just to summarize,

28:15 uh,

28:15 summarize these,

28:16 uh,

28:17 these results,

28:18 uh,

28:18 we find that

28:19 minimum pensions can be a powerful tool to reduce gender inequality,

28:23 uh,

28:24 gender,

28:25 gender pension gaps,

28:26 uh,

28:27 with an effect that is concentrated

28:29 on,

28:29 um,

28:30 on low schooling women and homemakers.

28:32 Um,

28:33 so

28:34 one thing that we look also a little bit in the paper is that,

28:37 um,

28:38 you know,

28:38 when you think about distributional effects of,

28:40 of this type of reform

28:42 or even disincentive,

28:43 uh,

28:43 uh,

28:45 uh,

28:45 effects,

28:46 um,

28:46 you have to take into account the household structure and the fact that.

28:49 Um,

28:50 the women with the least pension rights in

28:52 Chile and in many countries are homemakers who,

28:55 who just never contribute,

28:57 but they might be married with middle class or even high earning husbands.

29:00 So in terms of,

29:01 of,

29:01 uh,

29:01 distributional effects

29:03 and,

29:03 and disincentives,

29:04 you,

29:04 you need to,

29:05 to take that into account.

29:07 Um,

29:08 and we also find that the retirement age has

29:10 a first order effect on the gender pension gap,

29:13 um,

29:14 but that is concentrated on high schooling and,

29:16 and single women,

29:18 um,

29:18 and here

29:19 also you kind of wanna be careful because,

29:22 um,

29:23 essentially if,

29:23 if you're,

29:24 um,

29:25 if you're considering,

29:25 uh,

29:26 a rational agent,

29:27 right,

29:27 what you're doing is you're taking options away from that person,

29:30 right?

29:31 You're forcing them,

29:32 uh,

29:32 to,

29:33 to retire you're,

29:34 you're not allowing them to retire early.

29:37 And so that's not going to be a welfare enhancing for women and in fact that might be

29:41 part of why

29:42 this reform hasn't made it to to uh or this feature hasn't made it to the reforms,

29:47 uh,

29:47 even though it has been considered.

29:51 Um

29:53 OK.

29:53 And so,

29:54 uh,

29:55 now I would like to,

29:56 to,

29:57 to move on to the 3rd

29:58 set of results and the 3rd paper.

30:01 Um,

30:02 so,

30:02 um,

30:03 as I mentioned,

30:04 this is joint work with,

30:05 with Priyanka Kant,

30:05 who's,

30:06 uh,

30:06 who's working on health in the MENA region,

30:08 uh,

30:09 now.

30:10 And uh we're moving to the context of Pakistan and what we want to talk about is

30:15 um

30:16 um is about informal sector pensions.

30:20 Um,

30:21 but

30:21 you know,

30:22 backtracking a little bit,

30:23 um,

30:24 you know,

30:25 if we look at,

30:25 at,

30:26 um,

30:26 uh,

30:27 at,

30:27 uh,

30:28 many developing countries,

30:30 um,

30:30 or

30:31 let me start actually with,

30:32 with,

30:32 uh,

30:33 you know,

30:33 the recent,

30:34 um,

30:35 social protection and jobs compass

30:37 which uh sets as a,

30:38 as a long term goal universal

30:40 social protection.

30:42 But

30:42 in many countries what's going on at the moment is that you have

30:45 the poor who are targeted by social assistance

30:48 and uh formal workers who are covered by social insurance programs

30:52 and then uh in the middle you have the non-poor informal segments

30:57 um so in a in a

30:59 in a report that I co-authored with with Medis who

31:01 uh Medis G who's here in Himani uh Jain.

31:05 We take data from 6 countries in Africa and we,

31:09 we map out the segment and,

31:10 and look a little bit at at their characteristics

31:13 and,

31:14 um,

31:15 you know,

31:15 in,

31:15 in these,

31:16 in these countries,

31:17 um,

31:18 the non-poor informal segment is more than 50%

31:21 of,

31:21 of the population.

31:23 And we also find that it spans,

31:25 uh,

31:26 uh,

31:26 quintiles,

31:27 uh,

31:28 including the rich richest quintiles,

31:29 which is interesting.

31:31 Um,

31:32 we can also anticipate,

31:34 um,

31:35 uh,

31:35 as poverty,

31:36 uh,

31:37 rates decline and

31:39 formality doesn't look,

31:40 uh,

31:41 to decline or is stagnating,

31:43 that the segment might,

31:44 uh,

31:45 increase,

31:45 uh,

31:46 over time.

31:48 Um,

31:49 so,

31:50 um,

31:51 a growing number of countries are trying to,

31:53 um,

31:54 target this segment,

31:55 uh,

31:55 by introducing

31:57 voluntary pension schemes that are,

31:58 are designed specifically for,

32:00 for the informal sector.

32:03 However,

32:03 despite a few,

32:04 uh,

32:05 some promising,

32:06 um,

32:07 promising cases,

32:08 um,

32:09 most of these schemes have struggled to attract participants.

32:12 Um,

32:13 so

32:14 we can come up with a lot of reasons for why that might be the case.

32:17 Um,

32:18 it might be that,

32:19 uh,

32:19 workers are not

32:21 aware enough of these schemes,

32:23 uh,

32:23 that they don't understand them because of financial literacy.

32:26 It could be that they don't trust

32:28 the scheme and don't want to store their money in there.

32:31 Uh,

32:31 there might be a lot of behavioral obstacles to,

32:34 uh,

32:34 to registration and participation,

32:37 um,

32:37 but

32:38 the explanation that,

32:39 um,

32:40 we wanna

32:41 start,

32:41 um,

32:42 looking at here is maybe something more fundamental is maybe workers,

32:47 uh,

32:47 or informal workers just don't have the ability to

32:49 save and particularly to save for the long term.

32:52 Right,

32:53 it could be that they just are not,

32:54 uh,

32:55 they have low disposable incomes,

32:57 or they are,

32:58 they have uh their exposure to risks and their

33:00 liquidity needs are too high to um afford to,

33:04 to store money in a,

33:05 in a pension account.

33:08 So,

33:09 um,

33:09 this ability to save is,

33:11 is something,

33:11 is a concept that is very hard to,

33:13 to get to.

33:14 So

33:15 perhaps the first step,

33:16 uh,

33:16 um,

33:17 is,

33:17 uh,

33:18 what we're doing in this paper,

33:19 which is to just measure

33:20 what wealth accumulation looks like in the absence of a pension scheme.

33:25 And that

33:26 will

33:26 give you a first indication of the ability and desire

33:29 of,

33:29 of households to set,

33:31 uh,

33:31 money aside for the long term.

33:35 Um,

33:35 to do that,

33:36 we go to Pakistan and the reason is because,

33:39 um,

33:39 um,

33:40 there's a household survey in Pakistan that

33:42 asks consistent,

33:44 uh,

33:44 wealth questions,

33:45 uh,

33:46 over more than over 18 years.

33:48 Um,

33:49 and let me tell you,

33:50 I mean,

33:50 it's a well known problem,

33:51 but let me tell you why you need that much data.

33:54 So if you look at wealth accumulation

33:57 or in a given year by comparing younger,

34:00 uh,

34:00 household with a younger head

34:01 and households with a,

34:02 with an older head,

34:04 um,

34:05 you find some,

34:06 a profile that is relatively flat.

34:08 So that's the dashed line,

34:09 um,

34:09 in the graph on the right.

34:12 Um,

34:12 so this,

34:13 this is,

34:13 uh,

34:14 household net worth,

34:15 um,

34:15 in 2010,

34:17 right,

34:17 um,

34:17 by the age of the household head.

34:20 Um,

34:20 if you have data from

34:22 many surveys over time,

34:24 you can now track cohorts,

34:26 and,

34:26 uh,

34:26 you can see,

34:28 for example,

34:28 uh,

34:29 here in the,

34:29 with the 1960 birth cohort,

34:32 that the profile suddenly looks a lot steeper.

34:36 Um,

34:37 and so when we actually,

34:38 um,

34:39 decompose,

34:40 uh,

34:41 uh,

34:41 and,

34:41 and isolate the,

34:42 the effect of,

34:43 of,

34:43 uh,

34:44 age of the household head on net worth,

34:46 we find

34:47 that between the ages of 625 and 65,

34:50 a household accumulates the equivalent of

34:53 4 years' worth of

34:55 consumption.

34:55 So

34:56 the idea being that with that,

34:58 uh,

34:58 uh,

34:58 wealth accumulation,

35:00 if they sell everything,

35:01 they can survive for 4 years,

35:03 just to give you the,

35:03 the,

35:04 the thought experiment here.

35:08 And then we,

35:10 um,

35:10 we dig a little bit more into what type of wealth is accumulated and find that,

35:15 um,

35:16 uh,

35:16 almost all of that accumulation happens in housing,

35:19 so 80% of it,

35:21 and by contracts,

35:22 for example,

35:23 land is a sizable fraction of the portfolio,

35:25 but it doesn't

35:26 move over the,

35:27 uh,

35:27 with age that much.

35:30 And other types of uh

35:32 um instruments that we measure,

35:35 uh,

35:35 durables,

35:36 uh,

35:36 businesses,

35:37 uh,

35:37 financial wealth,

35:38 loans

35:39 that

35:40 um changes,

35:40 but the amounts are,

35:42 are dwarfed by,

35:43 by housing.

35:47 Um,

35:48 we also look at,

35:49 um,

35:50 accumulation rates,

35:51 uh,

35:52 over the consumption distribution.

35:54 And um

35:56 and what you see here is that uh the the rate of accumulation is actually very similar

36:02 across consumption quintiles.

36:04 uh you see that the first quintile accumulates a little bit slower than the others,

36:08 but other than that,

36:09 um,

36:10 the rates are similar.

36:11 Now,

36:12 um,

36:12 of course this is relative to household consumption,

36:15 right?

36:15 So clearly,

36:16 uh,

36:17 wealthier households are going to accumulate.

36:19 Um,

36:19 more wealth,

36:20 but relative to,

36:22 uh,

36:22 their level of consumption,

36:23 the,

36:24 the accumulation looks quite similar.

36:31 OK,

36:31 so,

36:32 uh,

36:32 to,

36:32 to conclude on this,

36:34 um,

36:34 you know,

36:34 this is a descriptive paper,

36:37 uh,

36:37 we just wanted,

36:38 wanted to put a number on,

36:39 on this question of,

36:40 you know,

36:41 of wealth accumulation

36:43 when there's no pension system,

36:44 and we find that,

36:45 uh,

36:46 accumulation is sizable,

36:48 uh,

36:48 about 4.2 years' worth of household consumption,

36:51 uh,

36:52 over,

36:52 over 40 years.

36:54 Uh,

36:54 we also find that that the rate of accumulation,

36:57 um,

36:57 gets faster after,

36:58 after age 40.

37:01 We find that most of the accumulation involves housing wealth,

37:04 which is interesting because

37:06 it kind of gives you also uh a sense of the need for liquidity,

37:09 right?

37:09 Like we,

37:10 we,

37:11 it indicates that households are able and willing to

37:14 save in a relatively illiquid form so housing is less

37:17 illiquid than let's say a pension account,

37:19 but

37:20 um it's still a relatively illiquid asset.

37:23 Um,

37:24 and we find that accumulation,

37:26 um,

37:27 looks,

37:27 um,

37:29 looks similar in relative terms between the poor and the rich,

37:32 which may be important because,

37:33 uh,

37:34 the question is,

37:34 do you want to target,

37:36 let's say those informal pension schemes

37:38 only to the wealthier parts of the informal sector or to all of the informal sector,

37:43 uh,

37:43 so this can,

37:44 can,

37:44 uh,

37:44 can,

37:45 um,

37:46 speak to,

37:46 to,

37:47 to that,

37:47 uh,

37:47 that problem as well.

37:51 All right,

37:51 so this gives me,

37:52 I think I still have a bit of time to,

37:55 to,

37:56 to give you,

37:56 um,

37:57 uh

37:58 some final remarks on,

37:59 on things that I think we don't know as much about

38:03 and that,

38:03 um,

38:03 I hope to,

38:04 to,

38:05 to be thinking about in the future.

38:07 Um,

38:08 so going back to this

38:10 non-poor informal segment,

38:12 um,

38:13 you know,

38:14 there's been a lot of research since the 90s on the poor and

38:17 on their financial and economic lives and we've learned a lot about,

38:20 about that.

38:21 There's also some research on the near poor and on transitions uh from,

38:26 uh,

38:26 from being not poor to being poor,

38:28 but.

38:29 Um,

38:30 there's still a lot of things that we don't really know very well about,

38:33 uh,

38:34 the,

38:34 the non-poor informal workers,

38:36 and this is going to be key as we try to move towards universal

38:39 social protection because these are the people that we still are left to,

38:43 to be covered.

38:44 Um,

38:45 so for example things like transitions between formality

38:48 and informality over the life cycle are not,

38:50 uh,

38:50 typically well described in most countries.

38:53 Um,

38:54 we tend to think of the formal and informal sector in a static way as,

38:57 you know,

38:57 you have the formal ones and the informal ones,

38:59 but

39:00 to what extent are,

39:01 are there transitions,

39:02 um,

39:02 in Chile as we see there's a lot of switchers,

39:05 um,

39:06 so that's going to matter for pension design.

39:09 Even things like shock exposure,

39:11 need for liquidity,

39:12 ability to self-insure,

39:14 um,

39:14 are often described,

39:15 well described in the cross section,

39:17 but,

39:17 uh,

39:17 we don't know very well to what extent it's the same households who are,

39:22 uh,

39:22 chronically exposed to shocks or,

39:24 uh,

39:24 whether,

39:25 uh,

39:25 it's,

39:26 it's uh more widespread.

39:28 Um,

39:29 and then expanding on the results on,

39:31 on housing wealth,

39:32 um,

39:33 you know,

39:33 the question now becomes,

39:35 uh,

39:36 do

39:36 households save in housing

39:38 because it's optimal for them for whatever reason or because it's a lack of option.

39:42 And,

39:43 and the,

39:44 and therefore,

39:44 you know,

39:45 would it be beneficial to,

39:47 um,

39:48 uh,

39:48 fix that labor that uh

39:51 financial market failure and,

39:52 and give them another instrument where they can,

39:54 um,

39:55 put some of that,

39:56 uh,

39:56 of that wealth and would,

39:57 would they gain gain from that.

40:00 Um,

40:01 and another,

40:02 um,

40:03 another topic that,

40:04 um,

40:05 I've,

40:05 I've started looking at with,

40:06 uh,

40:07 with,

40:07 uh,

40:07 Kathleen Beagle at the,

40:08 at the research in the research group is

40:10 the question of professional organizations in the informal sector.

40:13 So we tend to think of the informal sector as,

40:15 um,

40:16 atomistic,

40:16 uh,

40:17 you know.

40:17 Uh,

40:17 um,

40:18 individual workers,

40:20 but in fact anecdotal evidence,

40:21 uh,

40:21 and,

40:22 and you know,

40:22 many,

40:22 uh,

40:23 practitioners that we talked to during our cross support,

40:25 uh,

40:26 have a lot of things to say about,

40:27 uh,

40:28 informal sector professional

40:30 organizations.

40:31 So,

40:31 um,

40:32 the informal sector is organized.

40:34 Uh,

40:35 but we don't know much,

40:37 uh,

40:37 at least in terms of systematic documented,

40:39 um,

40:40 hard evidence,

40:41 uh,

40:41 in terms of what,

40:42 what they do,

40:43 these professional organizations,

40:44 and to what extent they might be providing some

40:47 insurance that,

40:48 and we,

40:48 we would want to take that into account for.

40:50 Um,

40:51 for,

40:52 um,

40:52 in,

40:52 in,

40:52 in,

40:53 in,

40:53 in,

40:54 in our operations,

40:55 um,

40:56 and also to what extent can they be leveraged

40:59 to,

40:59 uh,

40:59 to help deliver social protection programs,

41:02 uh,

41:02 to the informal sector,

41:04 so they can help

41:05 identify beneficiaries,

41:06 uh,

41:07 communicate to beneficiaries,

41:08 uh,

41:09 help build trust in the schemes,

41:11 uh,

41:11 maybe even more directly collect contributions,

41:14 um,

41:14 so I think that there's a lot of things that we,

41:16 we,

41:16 we need to learn about,

41:18 about this.

41:20 Um,

41:20 and just to,

41:22 to finish,

41:23 um,

41:24 another

41:24 kind of set of,

41:25 of questions that,

41:26 that,

41:26 um,

41:27 that I'm interested in,

41:29 um,

41:30 so

41:31 as we know,

41:32 population aging is,

41:32 is accelerating

41:34 and,

41:34 um,

41:35 we kind of know theoretically what needs to happen in order for,

41:38 uh,

41:38 countries to prepare for that,

41:40 right?

41:40 So we,

41:40 we need a lot of investment so that

41:42 the,

41:43 the,

41:43 the,

41:44 we can finance,

41:45 uh,

41:45 the cost of,

41:46 of,

41:46 uh,

41:47 of aging.

41:48 Um,

41:48 but I think

41:49 what,

41:50 what we don't see as much about is to what extent,

41:53 um,

41:54 decisions,

41:54 uh,

41:55 household decisions are adjusting to population aging,

41:58 um,

41:59 at both their decision and in fact their perceptions,

42:01 to what extent households are aware

42:03 that they're living in,

42:04 in an aging world and anticipating that,

42:07 um,

42:08 you know,

42:08 our individuals,

42:09 um,

42:10 expecting to work and live longer than,

42:12 um,

42:13 than their parents.

42:14 Um,

42:15 are they expecting to receive lower transfers from relatives,

42:18 uh,

42:18 and networks in old age?

42:21 And are they preparing for that,

42:22 um,

42:23 as a result?

42:24 Are they saving more

42:25 or are they saving differently than,

42:27 than,

42:27 um,

42:27 their parents did?

42:29 Um,

42:29 so for example,

42:30 in Pakistan,

42:30 even though we focus on the,

42:31 the age dimension,

42:33 we also see that.

42:34 Um,

42:35 saving also,

42:36 um,

42:37 uh,

42:37 increases across cohorts.

42:38 So the more,

42:39 uh,

42:39 recent cohorts to reach 65 years old tend to,

42:42 to reach that age with,

42:44 um,

42:44 with more,

42:45 uh,

42:45 wealth than,

42:45 than the,

42:46 the previous ones,

42:48 um,

42:48 and same thing our households investing more in children

42:51 in,

42:52 um,

42:53 because they anticipate,

42:54 uh,

42:54 that they're going to need their children to,

42:57 to be wealthier in order to,

42:58 to support them.

42:59 Um,

43:00 and lastly,

43:01 our traditional care and co-residence arrangements,

43:04 same thing,

43:05 responding to population aging,

43:06 uh,

43:07 in any way?

43:09 All right,

43:09 so that's,

43:10 um,

43:11 that's

43:11 all I have for today.

43:12 I just want to thank you again and thank you,

43:14 Michel,

43:15 and uh don't forget to wish uh Michel a happy birthday on your way out.

43:26 Thank you,

43:26 Clement for that.

43:27 Before turning over to Mihal,

43:28 let me just remind you,

43:29 if you're watching online,

43:30 please just raise your hand or put a note in the chat that you have a question and I,

43:34 I'll call on you,

43:35 uh,

43:36 during the Q&A session.

43:37 So with that,

43:37 over to,

43:38 to you,

43:38 Mihal.

43:42 um

43:43 Uh hello.

43:44 Uh good morning,

43:45 uh,

43:46 everybody,

43:46 and good afternoon to those who joined online from,

43:49 from Eastern Hemisphere.

43:51 It's a great pleasure to be here.

43:53 It's something which people often say at the beginning,

43:55 and I say it,

43:56 but I have a reason to say that.

43:58 I wanted to report to you that I'm

43:59 actually quite often invited

44:02 to some.

44:02 Speaking engagements,

44:03 but very seldom

44:05 recently invited to such an exciting speaking engagement,

44:08 which is commenting on a research paper.

44:10 So I really value it

44:12 even in annual meetings and all the retreats,

44:14 it will stick out as a memorable experience for me

44:17 compared to all others.

44:18 So thank you very much for that.

44:20 It's really a treat.

44:21 Uh,

44:22 let me say a few words about,

44:24 uh,

44:24 first about the context as I see it and then about the

44:28 paper,

44:28 uh,

44:29 and while talking about the paper,

44:30 I may,

44:31 uh,

44:31 mix it with suggestions for further research and,

44:35 and policy implications.

44:37 So the context is that the topic is,

44:39 uh,

44:39 is tremendously important

44:41 and,

44:42 uh,

44:42 more important,

44:43 uh,

44:43 more important,

44:44 uh,

44:45 every year.

44:46 Uh,

44:46 and that happens because of,

44:48 uh,

44:49 persistence of informality in the world.

44:52 Uh,

44:52 in,

44:52 uh,

44:53 in our white paper a few years ago of who main author is in the room,

44:57 uh,

44:57 we documented persistence of informality

45:01 and we also,

45:02 uh,

45:02 pointed out to the fact that in countries that became very formal,

45:06 the traditional employment relationships are in decline,

45:10 meaning,

45:11 uh,

45:11 working relationship become more fluid.

45:14 Uh,

45:14 there,

45:15 there are many more transitions in the labor market.

45:18 Uh,

45:18 in other words,

45:19 they start resembling more

45:21 the informal labor markets which we know so well from,

45:25 uh,

45:25 low and low and middle income countries.

45:27 So we have a persistence of informality

45:30 in,

45:31 in the low and low,

45:32 uh,

45:32 low and middle income countries and

45:34 growing fluidity in,

45:35 uh,

45:35 in,

45:36 um,

45:37 in high income countries

45:39 and,

45:39 uh.

45:41 In that context,

45:43 covering

45:44 informal sector by social protection has

45:46 become a fundamentally important thing for us

45:49 in the context which Clement described as,

45:52 uh,

45:52 our aim of universal social protection,

45:54 which is an international aim,

45:56 uh,

45:56 we have the universal social protection 2030 movement of,

46:00 uh,

46:00 of which we are a part of.

46:03 And um

46:05 and the question is tremendously difficult

46:08 uh

46:09 we are thinking about the whole of social protection,

46:11 but

46:12 uh Clement focused on pensions here and

46:15 pensions are probably

46:17 the most important of all social protection schemes for the following reason

46:22 that.

46:23 As countries become richer,

46:26 we see what's sometimes called

46:28 SEC,

46:28 social protection angle curve,

46:31 which is that

46:32 they spend more on social protection.

46:35 As countries become richer as a percentage of their spending and percentage of GDP

46:39 and then within the package of what they spent on social protection,

46:43 they spend more on pensions than on other social protection programs.

46:47 Again,

46:47 comparatively speaking,

46:49 so these are social protection angle curves here.

46:51 So informality and pensions is important it is,

46:54 it is already today,

46:56 it is a topic of very fast growing importance.

47:01 So,

47:01 um,

47:03 um,

47:03 and then you heard from Clement about this

47:05 group of non-poor informal which indeed remains,

47:09 uh,

47:09 has,

47:10 uh,

47:10 is,

47:10 is,

47:11 is of particular concern because

47:13 when you build social safety nets,

47:15 last resort social.

47:16 Assistance programs,

47:17 they typically concentrated on rural poor

47:20 and when you think of

47:22 existing formal schemes,

47:23 they focus on those who are formal.

47:27 So non-poor informal are neither here nor there,

47:30 and that's why it is a group of concern.

47:33 Moving to uh to the paper

47:35 itself,

47:36 um,

47:37 I,

47:37 and the,

47:38 the,

47:38 the let me talk in maybe 44 parts about 44 findings as I see it,

47:43 so.

47:43 So the first part

47:45 about the contribution rates and informality,

47:49 I really like the framework of,

47:51 of Clement that

47:53 informal sector is not really

47:54 merely a residual labor market,

47:56 but it offers competitive earnings opportunities.

48:00 uh,

48:00 and then in that context increase in contribution rates,

48:04 uh,

48:04 indeed increases informality and therefore reduces tax.

48:08 Revenue,

48:09 uh,

48:09 I think that,

48:10 um,

48:10 especially

48:11 with those with lower contribution densities,

48:14 I,

48:14 I,

48:14 I think

48:15 a method used is quite appropriate because we

48:18 have this method of simulated moments which,

48:21 as I understand it,

48:21 is if you have really

48:24 voluminous data

48:25 from various sources and then you had,

48:27 you had the longitudinal plus administrative,

48:30 then

48:31 this is a

48:32 fantastic way of

48:33 doing some kind of indirect inference and

48:37 And uh when it comes to uh findings,

48:40 which in a way would be expected,

48:42 although they are quite strong here,

48:45 um I think there are certain questions that come to

48:48 mind for both further researcher in terms of policy implications.

48:53 And the key question which is in my mind is

48:56 as follows.

48:56 So if

48:58 Uh,

48:59 can,

48:59 can,

49:00 can,

49:02 let me also add something which I should have added up front.

49:05 Any sort of

49:06 coverage of the

49:08 by social protection of the informal sector

49:11 would

49:11 probably have to take

49:13 a form of a combination of contributory and non-contributory.

49:17 A contributory being exactly what you examine Clement,

49:20 non-contributory being general revenue,

49:22 social assistance,

49:23 subsidization of contributions or whatever form works.

49:26 So

49:26 we are talking about the world when this combination be uh needed.

49:30 But

49:31 the key question which needs to be added first,

49:33 how much of that can be done by contributions,

49:35 right?

49:36 And

49:37 the question that comes to my mind reading your

49:39 paper is if countries were to take action to identify

49:43 right contribution rates,

49:45 not too high and not too low.

49:47 So that they would have a desired effect.

49:51 What would they be

49:52 and how would they relate to

49:55 both

49:56 voluntary savings

49:58 and optimal savings

49:59 and voluntary and optimal are not the same as we discussed a few,

50:03 a few years,

50:04 a few days ago

50:05 because of myopia,

50:07 workers may not workers,

50:08 informal workers may under,

50:10 uh,

50:10 under uh under save,

50:13 uh,

50:13 um,

50:14 so,

50:14 so,

50:14 but the relationship between.

50:17 Voluntary savings,

50:17 optimal savings,

50:19 and the contribution rates which are going to be used

50:21 for contributory coverage of the informal sector is phenomenally important,

50:27 important

50:28 question to answer,

50:30 uh,

50:30 trying to answer a bigger question.

50:31 Will the countries be able to

50:34 increase coverage

50:35 by contributory schemes in the informal sector without

50:38 formalization and without being able to impose,

50:41 to impose a mandate in in a in a hard way.

50:44 Um,

50:45 I like the,

50:46 the,

50:47 the second part of,

50:48 of the paper and the issue of minimum pension,

50:52 uh,

50:52 and the Clobex story.

50:54 Um,

50:55 I think it is,

50:56 uh,

50:57 actually

50:58 the fact that you identified 60,

51:00 I call it benefit withdrawal rate,

51:02 uh,

51:03 what you call loback,

51:04 but

51:05 you identified like 60% being the best

51:08 compared to other,

51:09 which I,

51:09 I reject,

51:10 say forget about 100 or 0,

51:12 right?

51:12 So we are talking where are those ranges,

51:15 and,

51:15 and you found 60% is quite intuitive,

51:18 um,

51:18 uh,

51:19 and I think the importance of um.

51:22 Uh,

51:22 of,

51:23 uh,

51:23 that

51:24 finding,

51:25 uh,

51:25 is,

51:26 is,

51:26 uh,

51:27 the importance of that

51:29 direction of,

51:29 of research is,

51:30 is tremendous because,

51:32 um,

51:33 because if we

51:34 need to

51:35 cover the informal sector by the combination

51:37 again we will be in the way to what

51:39 are the,

51:39 what's the right contribution.

51:41 Uh,

51:42 rate,

51:43 but then because we are talking about the whole of social protection,

51:47 all those who contribute are in the way in which they also,

51:50 some of them receive social assistance,

51:52 not contributory.

51:53 But then

51:54 you want to

51:55 withdraw social assistance if the level of income goes up,

51:58 right?

51:58 So the question is what should be the

52:00 benefit withdrawal rate to retain incentive to,

52:03 to first of all,

52:04 to,

52:04 to,

52:05 to work and then to contribute,

52:06 and this benefit withdrawal rate or the clawback rate,

52:08 it's not a clawback if it's maybe in the context of social assistance,

52:13 but I think this is a hugely important topic

52:16 and in fact both your findings,

52:18 I think we are highly relevant in.

52:20 The context of the

52:21 work done in luck where I think Truman,

52:23 you are leading it and I know Clement,

52:25 you are in the team.

52:26 I saw

52:26 the team we are discussing it next Tuesday

52:29 where the whole debate is about um

52:33 it's called formalization but I call it

52:35 uh

52:36 coverage

52:37 even without formalization how to

52:39 expand coverage to cover informal sector

52:42 which could be called

52:43 that if they are covered and they are formal,

52:44 right?

52:44 But

52:45 there are some fascinating parts.

52:46 Of that analysis.

52:47 One of them is the work on the formalization tax rate,

52:52 which is the proportion of the change in earnings

52:56 and which a worker loses when it moves into a formality which will be taxed away from,

53:01 uh from him or her.

53:02 I think that,

53:03 that your work on,

53:05 on,

53:05 on this has enormous relevance for that part of the work on the,

53:09 on the lab study.

53:09 So I think there is a lot of

53:11 uh potential,

53:12 potential there.

53:14 Uh.

53:16 Your gender gap findings are

53:19 actually the cleanest of all,

53:21 because to me they are both intuitive.

53:23 They are very solid,

53:24 and in so many policy dialogues we are making exactly those points,

53:29 maybe more about equalization of retirement age

53:31 and how women would benefit from equalization,

53:34 and then minimum wage point very well found in Chile.

53:37 So is often made.

53:39 So here we are in a situation where I think it's a little disagreement

53:43 among researchers,

53:44 among those who work on policies.

53:45 However,

53:45 there are big political obstacles.

53:47 In fact,

53:49 in fact,

53:49 uh,

53:50 speaking strictly,

53:51 uh,

53:52 the retirement age should probably depend on life expectancy,

53:56 uh,

53:56 uh,

53:56 at the age of retirement,

53:57 which would mean it should be higher for women.

54:00 Uh,

54:00 but I guess that that territory has not been entered in the discussion,

54:04 but maybe it should as a starting point to broader,

54:06 uh,

54:06 to broader,

54:07 uh,

54:07 uh,

54:08 to broaden the scope of discussion to arrive at reasonable discussion,

54:11 but I really like those,

54:12 uh,

54:13 findings.

54:14 And then when it comes to Euro-Pakistan analysis,

54:17 which is also something which I think is,

54:19 is,

54:20 is,

54:20 is,

54:20 is extremely interesting.

54:22 Uh,

54:23 it does,

54:24 uh,

54:24 and

54:24 I,

54:24 I will not repeat the findings,

54:26 but it poses,

54:27 it,

54:27 it poses actually a very interesting question

54:29 which,

54:30 uh,

54:30 I,

54:30 I myself thought very little about it,

54:33 which is about

54:35 that you have households who don't want to participate,

54:38 but they do have sizable savings

54:40 primarily in the form of,

54:42 uh,

54:42 savings in housing.

54:44 And there's a lot of questions why it happens,

54:49 what is special about that type of asset,

54:52 but there is also a question

54:53 about

54:55 Uh,

54:56 about,

54:57 is there a way,

54:58 um,

54:59 is there a way to,

55:01 uh,

55:02 To use those savings in housing for the purpose of

55:06 um

55:07 for the purpose of

55:09 uh

55:10 retirement

55:11 or disability

55:13 and uh is there a role for public policy in doing here anything and I think this is

55:20 in my vastly under researched and it is a topic of

55:24 uh of big importance

55:26 because we always always.

55:27 In policymaking,

55:28 we need to look at the whole picture,

55:30 not only at pensions.

55:31 We need to look at the whole of social protection,

55:33 but then we need to add a whole of savings.

55:35 So it's always,

55:36 it's like,

55:36 yeah,

55:37 like an onion when you have those,

55:38 uh,

55:38 those,

55:39 uh,

55:39 those,

55:40 uh,

55:40 layers.

55:41 And in fact,

55:43 this,

55:43 this,

55:43 your,

55:44 your,

55:44 your last set of findings uh related to Pakistan

55:48 it brings me to,

55:49 to,

55:49 to,

55:50 to,

55:50 to pointing out to the fact that

55:53 the work on it,

55:54 on the informality in the context of social.

55:57 Protection

55:58 we already talked about how important it is,

56:00 but let me offer a few strands of what is actually happening.

56:04 Medicine in the room is very much leading big strands of that of that work,

56:09 which is that

56:11 if.

56:13 I already said first let's think about what could be covered by contributory

56:17 and then obviously it's a question

56:19 maybe we should subsidize it and where should be social assistance but

56:22 starting with contributory,

56:24 the big issue is what is the role

56:27 of.

56:30 Behavioral in others,

56:32 what is the role of defaults to be set properly?

56:36 What is the role of bundling services like you bundle.

56:42 Old age security

56:44 with disability security and health insurance together

56:48 which may be more

56:51 that can may make informal sector workers more

56:53 willing to contribute if this is uh bundled

56:56 and

56:58 and

56:59 how you

57:00 allow for short term access along with long term access

57:04 as

57:05 Retirement pension savings need to be probably for that group accessible

57:09 in the case of other life changing events adversely affecting income.

57:14 So there is a huge set of,

57:15 I think,

57:16 research that needs to be done in,

57:18 in,

57:18 in this context.

57:20 What,

57:20 what is being done in our global practice,

57:23 and I already mentioned is here that we are carrying surveys

57:26 in several countries to.

57:29 What you came out pointed out how it's difficult to understand,

57:32 which is so-called willingness to save

57:34 our ability to save,

57:36 um,

57:36 so,

57:36 so that we hope that from those surveys we'll

57:38 be able to figure out the willingness to save

57:40 and that would contribute to our knowledge base which

57:43 would ultimately may inform thinking about how to cover

57:47 informal sector worker with contributory schemes or

57:49 in what part to cover uh cover them

57:52 um and then based on that

57:54 what.

57:55 Subsidization covered by general revenue

57:58 would be appropriate and what what rate would be appropriate

58:02 and this is something this is considered by some countries that combination

58:05 and I think this is where the sort of nerve of

58:08 further research should be.

58:10 Um,

58:10 and it should be there again,

58:11 I repeat,

58:12 because we cannot

58:13 at all rely on formalization.

58:15 It's,

58:16 it ain't going to happen.

58:17 It hasn't happened for

58:19 40 years.

58:20 I love saying India was 90% informal in 1948.

58:23 It's 90% informal now.

58:25 Nothing changed there and

58:27 trying to think in terms of formalizing would be like the definition of insanity,

58:31 doing the same again,

58:32 hoping for a different result,

58:33 right?

58:34 So you need a different approach.

58:35 The approach we are using is let's cover them.

58:39 Without formalizing or if in the process they formalize this even better,

58:42 but the formalization should not be a center of

58:46 a central point of reference because that may immediately lead us to

58:51 wrong policy conclusions just by the way that we call it formalization.

58:56 Let's let's tax them.

58:57 Let's make it mandatory.

58:58 Let's impose the law,

58:59 which,

58:59 you know,

59:00 you can do all kinds of wrong things the moment you start,

59:02 the moment you think how,

59:03 how using behavioral knowledges,

59:05 using soft mechanisms,

59:06 encourage them becoming more formal with

59:09 respect to social security contributions,

59:11 because yet another issue is.

59:13 Definitions of informality you can be formal with respect to

59:16 taxes but not social security contributions or the other way around

59:19 or formal with real estate taxes but not other taxes.

59:23 I think there is a lot of work I've seen

59:25 part of that in showing those different shades of gray.

59:30 Let me conclude by

59:31 Yeah,

59:31 uh,

59:32 and one thing which you said,

59:33 which,

59:34 which I think

59:35 I seldom hear and I liked what I heard is you

59:38 pointed out the role of professional organizations in the informal sector,

59:43 and I

59:44 seldom hear discussions on that.

59:46 Maybe I,

59:46 I just don't read the right things,

59:48 but obviously intuitively speaking,

59:50 I could think of

59:51 organizations like SEWA in India

59:54 and several other self-group.

59:57 Of women,

59:58 which played an enormously important role in the development of labor

1:00:02 force participation there and could potentially play a very important role

1:00:06 in expanding coverage of contributory schemes to the formal sector.

1:00:11 So this is something which I'm happy you and Kalin

1:00:14 are going to look at because it's tremendously important.

1:00:16 So congratulations and thank you very much.

1:00:17 The right discussions at the right time

1:00:19 and the importance of that will be growing every year.

1:00:22 Thank you.

1:00:23 Thanks Michel for that.

1:00:24 That's uh both the reflections also the,

1:00:27 the very clear message about the role and the informality is here to stay,

1:00:32 and we need to think about sort of policies in that context,

1:00:35 um,

1:00:35 before turning it back to,

1:00:37 to,

1:00:37 to Clement,

1:00:38 uh,

1:00:38 we have one hand raised,

1:00:40 so I'm just gonna ask Gustavo,

1:00:42 oh yeah,

1:00:42 Gustavo,

1:00:43 are you there?

1:00:44 Um,

1:00:45 and then I'll turn it back to Clement to res react maybe to,

1:00:47 to,

1:00:48 to Mikhael's point and maybe Gustavo's point,

1:00:50 uh,

1:00:50 question as well.

1:00:51 So go ahead,

1:00:52 Gustavo.

1:00:54 Uh,

1:00:55 thank you very much,

1:00:56 Ion,

1:00:56 and thank you,

1:00:57 uh,

1:00:57 Clement,

1:00:58 for an excellent,

1:00:59 uh,

1:01:00 presentation but also

1:01:02 for an excellent collaboration with,

1:01:04 uh,

1:01:04 with,

1:01:05 uh,

1:01:05 our teams on several,

1:01:07 uh,

1:01:07 um,

1:01:08 fronts of,

1:01:08 uh,

1:01:09 the work on

1:01:10 informality.

1:01:11 I think this is,

1:01:12 uh,

1:01:12 this is,

1:01:13 um,

1:01:13 uh,

1:01:14 a really an excellent

1:01:16 summary of the analytical work,

1:01:18 uh.

1:01:19 Uh,

1:01:19 and,

1:01:20 and,

1:01:20 and the link with operational work that,

1:01:22 that,

1:01:22 uh,

1:01:23 you have been doing.

1:01:24 Let me,

1:01:24 let me,

1:01:25 uh,

1:01:26 make a couple of,

1:01:27 uh,

1:01:28 comments for your reactions.

1:01:30 So you mentioned in the first part of your presentation,

1:01:34 um.

1:01:35 The uh the,

1:01:37 the possible

1:01:39 er um implications of uh regulations on uh on informality.

1:01:44 And you focus mostly

1:01:47 on uh contribution rates

1:01:50 uh when when you mentioned other um

1:01:53 um regulations such as minimum pensions

1:01:57 you highlighted the importance to reduce inequalities

1:02:01 but there are some uh

1:02:03 also some concerns that minimum pensions and the.

1:02:06 Expansion of uh social pensions

1:02:09 can be a disincentive

1:02:11 to formalize and I would like to get your your reactions on on on this.

1:02:18 There's one point you didn't mention.

1:02:20 You talked about the uh the contribution density

1:02:24 and you didn't mention the best thing,

1:02:26 uh,

1:02:26 that,

1:02:26 that the,

1:02:27 the,

1:02:27 the possible.

1:02:29 This incentive

1:02:31 of um

1:02:32 investing periods of contributions so we have

1:02:36 for example in Latin America countries where

1:02:39 you need to have 10 years of contribution to get a

1:02:41 pension and you have countries where you need to have 30 years

1:02:45 of contributions to get a

1:02:47 pension.

1:02:47 Do you think these may potentially have.

1:02:51 Anything uh to do with um

1:02:54 the disincentives

1:02:56 and

1:02:57 to formalize when people think it's going to be unrealistic or not very easy to

1:03:03 um.

1:03:04 Uh,

1:03:05 accumulate 30 years of contribution,

1:03:07 but at the same time,

1:03:08 uh,

1:03:09 they may have easier access

1:03:11 to minimum pensions or universal,

1:03:14 uh,

1:03:14 basic pensions

1:03:15 and,

1:03:17 uh,

1:03:17 my second

1:03:18 question is,

1:03:19 uh,

1:03:19 on your important contribution to understanding the savings potential

1:03:24 of the informal sector and I think this is,

1:03:26 uh,

1:03:27 this is really

1:03:28 a very fundamental piece

1:03:30 and something that,

1:03:31 uh,

1:03:32 that has changed,

1:03:33 uh,

1:03:33 our.

1:03:34 Uh,

1:03:35 vision of uh of uh

1:03:38 what solutions may work for the informal sector.

1:03:41 So there's a lot of evidence and you accumulated more evidence

1:03:45 you produce more evidence

1:03:47 that that there is potential of savings

1:03:51 in the informal sector.

1:03:53 So,

1:03:54 um,

1:03:54 the question is,

1:03:56 what else do we need to do

1:03:58 to try to,

1:03:59 uh,

1:04:00 use that potential or to drive that potential or to encourage

1:04:04 that potential to long term savings versus other forms of savings.

1:04:08 Thank you.

1:04:12 Thanks,

1:04:12 Gustavo,

1:04:13 um.

1:04:14 Maybe Clement,

1:04:15 if you want to react to sort of that collection of of of reactions and

1:04:19 take it from there.

1:04:21 Yes,

1:04:21 um,

1:04:22 thank you,

1:04:22 um,

1:04:23 both,

1:04:23 uh,

1:04:23 Gustavo and,

1:04:24 and Michel for these comments.

1:04:26 Um.

1:04:28 So

1:04:30 maybe taking the,

1:04:31 the comments of Mijal first,

1:04:33 I think,

1:04:33 uh,

1:04:34 uh,

1:04:34 a broad,

1:04:34 um,

1:04:35 a broad thought that I have on,

1:04:37 you know,

1:04:37 choosing the optimal

1:04:39 contribution rate and the optimal clawback rate and that's

1:04:41 very obvious when you look at Chile is that,

1:04:44 um,

1:04:45 political economy considerations,

1:04:47 uh,

1:04:47 often

1:04:48 trump whatever,

1:04:49 you know,

1:04:49 optimal rate you might find by doing the type of exercise that,

1:04:52 that,

1:04:52 uh,

1:04:52 that I've done.

1:04:53 Um,

1:04:54 as you can see,

1:04:54 you know,

1:04:54 in Chile,

1:04:55 the,

1:04:55 it's not even clear that the system will,

1:04:57 will survive,

1:04:58 uh,

1:04:58 in the long term.

1:04:59 So,

1:05:00 um,

1:05:01 that's why I tried to,

1:05:01 to mention a little,

1:05:02 a little bit in the talk,

1:05:03 but.

1:05:04 Um,

1:05:06 yeah,

1:05:06 when you,

1:05:06 when you,

1:05:07 when you choose the optimal contribution rate,

1:05:08 it might be that the,

1:05:09 the,

1:05:10 the,

1:05:10 the optimal rate that minimizes disincentives to formality is X.

1:05:16 So in fact,

1:05:16 I,

1:05:16 I think,

1:05:17 uh,

1:05:17 uh,

1:05:18 uh,

1:05:18 we do an exercise like that in,

1:05:19 in,

1:05:19 in,

1:05:20 uh,

1:05:20 I do an exercise like that in the paper,

1:05:21 but,

1:05:22 um,

1:05:23 it seems,

1:05:23 or

1:05:24 one thing that,

1:05:24 uh,

1:05:25 that I've became convinced on,

1:05:26 uh,

1:05:27 of,

1:05:27 um,

1:05:28 uh,

1:05:28 more recently is that.

1:05:29 Um,

1:05:30 you kind of need a contribution rate,

1:05:32 or it has a value of also of a symbol,

1:05:34 and so you need a contribution rate that leads

1:05:36 to a pension that people can live on and,

1:05:38 you know,

1:05:39 even though,

1:05:40 you know,

1:05:40 in our,

1:05:41 in our model,

1:05:41 we could,

1:05:42 we would assume that people can,

1:05:43 can complement their,

1:05:45 their,

1:05:46 their,

1:05:46 um,

1:05:47 lower,

1:05:47 a lower contribution rate with uh voluntary savings,

1:05:50 but I think

1:05:51 the,

1:05:51 the,

1:05:52 the contribution rate of the,

1:05:53 of the system kind of needs to lead to.

1:05:56 To a,

1:05:56 a,

1:05:56 a pension that people can,

1:05:57 can,

1:05:58 can live on.

1:05:58 So I think,

1:05:59 and,

1:06:00 um,

1:06:01 um,

1:06:01 and after that you can,

1:06:02 you kind of wanna take into account that it might create some this some informality,

1:06:06 etc.

1:06:06 but,

1:06:07 but I think that's a,

1:06:08 a primary consideration

1:06:09 and likewise for the,

1:06:11 for the clawback rate kind of a similar point is that um.

1:06:15 You know,

1:06:15 in the paper we kind of see,

1:06:17 you know,

1:06:17 the exercise that we do is to say if the goal is to,

1:06:21 to,

1:06:22 uh,

1:06:23 sustain,

1:06:23 uh,

1:06:23 um,

1:06:24 uh,

1:06:24 consumption floor,

1:06:25 what's the cheapest

1:06:26 way to do it in terms of choosing that clawback rate,

1:06:29 and,

1:06:29 and we find that

1:06:30 you want a relatively aggressive clawback of 60% or even higher,

1:06:34 um,

1:06:35 but,

1:06:36 uh,

1:06:36 and,

1:06:37 and in Chile they,

1:06:37 they chose a much lower clawback rate.

1:06:40 Um,

1:06:41 and,

1:06:41 but part of that could be again political considerations,

1:06:44 right,

1:06:45 because you want to transfer some money to the,

1:06:47 to the middle class and you do that by having a relatively soft clawback rate so that,

1:06:51 uh,

1:06:51 even people with relatively

1:06:54 higher pensions are gonna get something from the government,

1:06:57 um,

1:06:57 so again,

1:06:58 I mean also in this case really uh political economy considerations are going to be,

1:07:02 to be very important to take into account.

1:07:04 Um,

1:07:07 Let me see.

1:07:08 So,

1:07:08 so in terms of,

1:07:09 um,

1:07:10 uh,

1:07:12 so you mentioned something that in terms of housing and whether,

1:07:15 um,

1:07:16 uh,

1:07:17 it can be used uh or whether that wealthy community housing should be or could be.

1:07:22 Um,

1:07:23 diverted into,

1:07:24 um,

1:07:25 into pension programs,

1:07:27 uh,

1:07:27 to be honest,

1:07:28 I think,

1:07:28 um,

1:07:29 you know,

1:07:29 this is a recent paper,

1:07:30 and we,

1:07:31 we,

1:07:31 we find that result,

1:07:32 but I'm not,

1:07:33 um,

1:07:33 you know,

1:07:33 a,

1:07:34 a real estate economist,

1:07:35 and so I think,

1:07:36 uh,

1:07:36 I still need to learn a bit more about,

1:07:38 um,

1:07:39 what is out there in terms of why people,

1:07:41 um,

1:07:42 saving housing.

1:07:43 I,

1:07:43 I don't know if,

1:07:44 if,

1:07:44 uh,

1:07:45 everything is out there,

1:07:45 but,

1:07:45 but certainly.

1:07:46 Um,

1:07:47 I,

1:07:47 I,

1:07:48 and perhaps we need to understand a little bit more why,

1:07:51 uh,

1:07:51 that's the case,

1:07:51 why,

1:07:52 you know,

1:07:52 what,

1:07:52 what,

1:07:53 what roles,

1:07:53 you know,

1:07:53 housing is a very specific type of assets,

1:07:55 and so what,

1:07:56 what's its role,

1:07:57 um,

1:07:58 so I think in the literature,

1:07:59 you know,

1:07:59 that there's the consumption value of housing that is taken into account,

1:08:02 but I don't know if the.

1:08:04 Uh,

1:08:04 maybe it can be used also to

1:08:06 facilitate intragenerational transfers where you know,

1:08:09 you,

1:08:10 you have a big house and maybe you can

1:08:12 reside with your children and in exchange for transfers.

1:08:15 So all of that,

1:08:16 um,

1:08:16 I think needs a little bit more research to be understood to

1:08:19 see whether in fact households are overinvesting in housing and because they,

1:08:23 that's the only safe,

1:08:25 uh,

1:08:25 instrument that they have or,

1:08:26 or because that's,

1:08:28 uh,

1:08:28 the,

1:08:28 the optimal thing they can do.

1:08:31 Um,

1:08:33 Uh,

1:08:34 and so I think,

1:08:35 I mean you,

1:08:36 you pointed out,

1:08:37 uh,

1:08:37 a lot of,

1:08:37 um,

1:08:38 of,

1:08:38 of,

1:08:38 um,

1:08:39 of open questions on in terms of behavioral nudges and,

1:08:42 and bundling and,

1:08:43 and,

1:08:43 uh,

1:08:44 um,

1:08:44 allowing for some short term access to these,

1:08:47 these pension schemes and whether that would,

1:08:49 would,

1:08:49 uh,

1:08:49 increase participation.

1:08:51 I agree,

1:08:51 so that's very,

1:08:52 uh,

1:08:52 very much,

1:08:52 uh.

1:08:53 Um,

1:08:55 uh,

1:08:55 the,

1:08:55 the frontier and,

1:08:56 and,

1:08:57 and should be,

1:08:57 should be researched.

1:08:58 Uh,

1:08:59 sometimes it's the difficulty with pensions is that it's hard to experiment,

1:09:02 right,

1:09:02 because once you start

1:09:04 implementing a pension system in a certain way,

1:09:06 you cannot really change it after,

1:09:07 after two years easily.

1:09:08 So that's,

1:09:09 that's really,

1:09:09 uh,

1:09:10 the

1:09:10 probably why,

1:09:11 why we don't know more about these,

1:09:12 these topics,

1:09:13 but,

1:09:13 um,

1:09:14 that's why,

1:09:15 you know,

1:09:15 you can try to,

1:09:16 to,

1:09:16 to get at these questions uh in in direct ways as we've tried to do it in,

1:09:19 in Pakistan.

1:09:21 Um,

1:09:22 and then,

1:09:22 uh,

1:09:23 moving on to Gustavo's remarks,

1:09:25 um.

1:09:27 On the,

1:09:27 the disincentives,

1:09:28 um,

1:09:29 uh,

1:09:30 about the minimum pension,

1:09:30 I think in the talk,

1:09:31 I talked,

1:09:32 uh,

1:09:33 you know,

1:09:33 I,

1:09:33 I talked mostly about the choice of that,

1:09:35 that clawback rate,

1:09:36 and what we find is that it doesn't,

1:09:38 in terms of formality,

1:09:39 it doesn't matter that much whether you claw back,

1:09:41 um,

1:09:41 uh,

1:09:42 more aggressively or or less aggressively,

1:09:44 uh,

1:09:45 but that's not the whole story because,

1:09:46 um,

1:09:47 you know,

1:09:48 the,

1:09:48 I think one first order impact of minimum pension is what

1:09:51 we call a wealth effect which is that once you start.

1:09:53 Uh,

1:09:54 transferring more money to people they're likely to retire earlier,

1:09:57 um,

1:09:57 and you know,

1:09:58 to an extent that's kind of what we want,

1:09:59 right?

1:10:00 Like the minimum,

1:10:01 the minimum pension is also there so that people don't have to work until they,

1:10:04 they are completely,

1:10:05 uh,

1:10:06 um,

1:10:06 disabled and,

1:10:07 and unproductive,

1:10:08 uh,

1:10:08 but it's,

1:10:09 it's to allow people to,

1:10:10 to to retire at a decent age.

1:10:12 So we also want to,

1:10:13 to keep that in mind that,

1:10:14 that,

1:10:14 that's also what we're trying to achieve by having a minimum pension.

1:10:17 Um,

1:10:19 but yeah,

1:10:19 so I,

1:10:19 I guess my,

1:10:20 the message of

1:10:21 my research so far is that,

1:10:22 you know,

1:10:23 the,

1:10:23 the,

1:10:23 the,

1:10:23 the,

1:10:24 the,

1:10:24 this implicit taxes consideration,

1:10:26 they don't matter as much as you might think.

1:10:29 Uh,

1:10:29 really,

1:10:30 what's going to matter is what level of,

1:10:32 of consumption you want to guarantee,

1:10:33 taking into account the fact that,

1:10:35 yeah,

1:10:35 that's going to allow people to retire earlier,

1:10:37 uh,

1:10:38 and,

1:10:38 and,

1:10:38 and maybe that's part of what we want to do.

1:10:41 Um,

1:10:42 vesting periods,

1:10:43 uh,

1:10:43 I haven't looked at that,

1:10:44 uh,

1:10:45 very much,

1:10:45 even though it's,

1:10:46 it's,

1:10:46 um,

1:10:47 uh,

1:10:47 it,

1:10:47 it,

1:10:47 it's very interesting.

1:10:48 I think Truman,

1:10:49 uh,

1:10:49 has,

1:10:49 has worked on that,

1:10:51 um,

1:10:51 um,

1:10:52 in my,

1:10:52 my reading of the Chilean cases because in Chile they

1:10:55 did have a vesting period for the minimum pension,

1:10:57 but.

1:10:57 Not that many people ended up right at that margin again,

1:11:01 because uh you have a lot of,

1:11:02 you have the people who work formally all the time and they will be

1:11:05 way past the vesting period and then people who never work formally and they're,

1:11:09 uh,

1:11:09 they,

1:11:10 they will never dream to,

1:11:11 to reach that vest vesting period.

1:11:13 So that reduces a little bit the margin that you're,

1:11:15 that is at play when you're considering these disincentives.

1:11:19 Um,

1:11:20 um,

1:11:22 and,

1:11:22 um.

1:11:24 And I think I forgot your last question,

1:11:25 so uh maybe you can remind me,

1:11:27 or,

1:11:27 or otherwise,

1:11:28 uh,

1:11:29 yeah,

1:11:29 I don't know if there's other questions.

1:11:31 We have 4 questions in the room,

1:11:32 so can I ask you to maybe just go quickly?

1:11:34 Um,

1:11:35 we have a little time,

1:11:36 but,

1:11:36 but,

1:11:36 um,

1:11:37 maybe start closest to me and then go around,

1:11:40 so.

1:11:42 Um,

1:11:43 hi.

1:11:43 Uh,

1:11:44 my name is Priyanka.

1:11:45 I'm with,

1:11:45 uh,

1:11:46 H&P.

1:11:46 I actually moved from SP last year,

1:11:48 and we're working,

1:11:49 um,

1:11:49 on social health insurance,

1:11:51 um,

1:11:52 and trying to understand,

1:11:53 you know,

1:11:53 some of these issues,

1:11:54 um.

1:11:56 With respect to willingness to pay,

1:11:57 with respect to,

1:11:58 you know,

1:11:59 compliance,

1:11:59 and I see a lot of similarities,

1:12:01 um.

1:12:02 In terms of,

1:12:02 you know,

1:12:03 pension pension contribution

1:12:05 being

1:12:06 You could use that as a proxy for compliance for estimating compliance,

1:12:10 uh,

1:12:10 minimum level of benefits,

1:12:12 um,

1:12:13 you know,

1:12:13 the hyperbolic discounting that you speak about so

1:12:16 one question is,

1:12:17 um,

1:12:18 whether you think we could use a similar framework for understanding,

1:12:22 um.

1:12:23 Some of these issues with respect to social health insurance and

1:12:27 um I guess the other question is

1:12:29 if if there is all this work happening on informality is is this also

1:12:33 something that's being looked at and if so I'd love to know more about it

1:12:38 thanks.

1:12:43 Thanks so much and happy birthday Mihal.

1:12:45 Um,

1:12:46 my name is Katriel.

1:12:47 Uh,

1:12:48 maybe this is less important in light of the political economy

1:12:51 considerations that you just highlighted,

1:12:52 Clement,

1:12:52 but,

1:12:53 um.

1:12:54 I'm curious how seriously.

1:12:57 How confident we can be in the estimate of the effect of

1:13:01 um contribution rates on

1:13:03 informality.

1:13:04 And I sort of see two subs questions

1:13:06 that this is,

1:13:07 I think,

1:13:07 especially,

1:13:08 I mean,

1:13:08 it's especially interesting whether that translates into

1:13:12 Um,

1:13:13 affect

1:13:13 the estimate,

1:13:14 the true effect in,

1:13:15 in poorer countries than Chile.

1:13:17 Um,

1:13:18 one sub question is like,

1:13:19 what are the parameters that are really driving that?

1:13:21 Is that the,

1:13:22 is that the rate of time discount,

1:13:24 sorry,

1:13:24 time preference,

1:13:25 um,

1:13:25 or is it the other kinds of preferences that people have,

1:13:28 um,

1:13:29 and then.

1:13:30 The

1:13:32 The second question is just the robustness of that kind of that structural

1:13:35 model.

1:13:36 Do you have,

1:13:37 uh,

1:13:37 has it,

1:13:38 does it make predictions,

1:13:39 is it,

1:13:40 has it made predictions that you've been able to

1:13:41 validate on sort of out of sample data?

1:13:44 Over.

1:13:49 So

1:13:49 my question follows uh a bit what Mihal um

1:13:54 uh

1:13:54 mentioned and I'm not sure to what extent this incorporated both in

1:13:57 in research and in in the global practice and in in governments

1:14:02 to what extent,

1:14:03 uh,

1:14:04 there is a uh it's when

1:14:06 I mean you describe the issue of pension,

1:14:08 but pension is usually bundled I mean it's associated

1:14:10 with other benefits that workers might have when they,

1:14:13 they join the workforce.

1:14:15 Including

1:14:16 maybe health insurance,

1:14:18 uh,

1:14:18 like,

1:14:19 um,

1:14:19 disability insurance,

1:14:20 etc.

1:14:20 and how much taxes they're,

1:14:22 they're going to be paying

1:14:23 and in and,

1:14:25 and so I mean the decision to work on or not work,

1:14:28 uh,

1:14:28 in the formal sector is not just great to the pension.

1:14:31 There is a lot of,

1:14:31 um,

1:14:32 things associated with that

1:14:33 and also,

1:14:34 uh,

1:14:34 from the government side,

1:14:36 uh,

1:14:36 whether all those benefits are provided or not will affect the.

1:14:39 Person's decision to work in the formal sector or not.

1:14:43 So my question is

1:14:44 to what extent both in the research and in the in in the practice

1:14:48 governments have taken that into account how

1:14:50 for example they are adjusting uh taxation,

1:14:53 uh,

1:14:54 income taxes or not income taxes,

1:14:56 uh,

1:14:57 that will affect the the decision of workers to both to work in the formal sector

1:15:01 but also,

1:15:02 uh,

1:15:03 how much money they will have to pay for the pensions that they might be providing.

1:15:09 Thanks,

1:15:10 Miller.

1:15:13 Sorry,

1:15:13 I'm,

1:15:13 I'm gonna call on Melis,

1:15:14 and then,

1:15:15 and then just after Melis I'm gonna call on Gonzalo who's got a hand raised,

1:15:19 and then I'll turn it back to the presenters and then we'll conclude.

1:15:21 So go ahead.

1:15:22 Thank you so much,

1:15:23 um,

1:15:23 and thanks so much,

1:15:24 uh,

1:15:24 Clement,

1:15:25 uh,

1:15:25 for the presentation and as well as for the collaboration.

1:15:29 I think,

1:15:29 uh,

1:15:29 with the work we've done together,

1:15:32 uh,

1:15:32 we managed to.

1:15:33 Uh,

1:15:34 operationalize,

1:15:35 um,

1:15:35 informal sector,

1:15:36 um,

1:15:37 savings,

1:15:38 um,

1:15:39 um,

1:15:39 or,

1:15:40 or pension scheme in a couple of countries,

1:15:42 um,

1:15:42 we are working

1:15:43 on another one currently,

1:15:45 so,

1:15:45 so I think,

1:15:46 uh,

1:15:46 that's really important and to

1:15:48 complement the work as,

1:15:49 as Mihal mentioned,

1:15:50 uh,

1:15:51 we are carrying out some surveys to inform design.

1:15:54 And I think the one in Pakistan um um maybe respond to some of

1:15:59 the questions that um that uh are put on the table in terms of

1:16:03 maybe why

1:16:05 um

1:16:05 you know informal sector

1:16:07 might be investing in housing

1:16:09 it's not a specific

1:16:10 question as to housing,

1:16:12 but you know there are

1:16:13 questions related to trust in terms of

1:16:16 where the informal sector um

1:16:18 would be actually uh willing to and and be comfortable uh to to put their um money in.

1:16:24 And I just also wanted to mention uh about the integrators right?

1:16:28 uh it's real because um these integrators these

1:16:31 um informal sector associations are being relied on

1:16:36 significantly in Rwanda

1:16:38 uh to expand coverage,

1:16:39 uh,

1:16:40 certainly in the MMA scheme in Kenya,

1:16:43 uh,

1:16:43 and,

1:16:43 uh.

1:16:44 There's another scheme just now which is called Habahaba in Kenya and they do also,

1:16:48 um,

1:16:49 rely on,

1:16:49 uh,

1:16:49 informal sector associations and and cooperatives

1:16:53 as um as integrators.

1:16:55 So,

1:16:55 so they have been successful,

1:16:57 uh,

1:16:57 in those areas where I think,

1:16:59 um,

1:17:00 in a way the informal workers are are organized.

1:17:03 Thank you.

1:17:05 Thanks,

1:17:05 Melis.

1:17:06 Uh,

1:17:07 Gonzalo,

1:17:07 are you still there?

1:17:10 Yes,

1:17:11 thank you very much.

1:17:11 Uh,

1:17:12 thanks for the presentation.

1:17:14 Uh,

1:17:14 my question is about the,

1:17:16 uh,

1:17:16 well I have two comments about the,

1:17:18 the,

1:17:18 the knowledge gaps.

1:17:19 Um,

1:17:20 one is,

1:17:21 uh,

1:17:21 in terms of the the nature of the informal

1:17:25 sector

1:17:26 that I think is also changing and we have now,

1:17:28 for example,

1:17:29 the,

1:17:29 the emergence of the gig economy.

1:17:31 So how we can,

1:17:32 um,

1:17:34 also include that as a,

1:17:35 as a feature in our,

1:17:36 in our research,

1:17:37 uh,

1:17:38 and.

1:17:39 If that open if that opens a new challenge or even gives us new opportunities

1:17:45 to um

1:17:47 uh

1:17:47 to

1:17:48 tap into to tap into the the possibilities of

1:17:51 using

1:17:52 the semi semi organized

1:17:55 uh gig gig uh

1:17:58 informal sector that is now

1:18:00 integrated into the gig economy because they're organized

1:18:03 through apps and this this type of work

1:18:05 so that is a new,

1:18:06 a new feature that gives a new opportunity.

1:18:09 And the second is the um

1:18:12 the 2nd,

1:18:12 the 2nd is uh

1:18:14 on the uh

1:18:17 On the possibility of tapping into other sources of of of contributions

1:18:23 from the informal sector,

1:18:25 in particular what we have been thinking about is the

1:18:28 informal sector or the informal income is not very visible,

1:18:32 but all the.

1:18:33 Informal

1:18:34 workers,

1:18:35 uh,

1:18:36 they have to consume at some point

1:18:38 and whether consumption will be more visible to tap into as a source of,

1:18:41 of contributions.

1:18:43 So that is another,

1:18:45 uh,

1:18:46 part of the research agenda that we can,

1:18:48 we can expand.

1:18:49 Thanks.

1:18:52 Thanks,

1:18:53 uh,

1:18:54 thanks,

1:18:54 Gonzalo.

1:18:55 So I'm gonna let Clement maybe react and maybe conclude from your side,

1:18:58 and then I'll close with,

1:18:59 with Mikhael to close the session.

1:19:00 Thanks.

1:19:01 Um,

1:19:02 thank you.

1:19:02 Yes,

1:19:03 um,

1:19:04 all right,

1:19:04 so Priyanka,

1:19:05 yes,

1:19:05 I,

1:19:05 I agree,

1:19:06 Priyanka,

1:19:06 that,

1:19:06 that a lot of the same questions,

1:19:08 um.

1:19:09 Or,

1:19:09 um,

1:19:10 that relates to participation in pen in pension schemes will be relevant for,

1:19:14 for health insurance schemes even though,

1:19:16 uh,

1:19:17 often health insurance,

1:19:18 you know,

1:19:18 provides more immediate benefits so there's less of this question of present bias,

1:19:21 etc.

1:19:23 uh,

1:19:23 but I think in both cases,

1:19:25 as I said,

1:19:26 and there's a lot of things we need to,

1:19:27 to learn or document about

1:19:29 the these non-poor informal sector that we're trying to incorporate

1:19:33 into the schemes that will be useful for both,

1:19:35 uh,

1:19:36 uh,

1:19:36 for both,

1:19:37 um.

1:19:39 Um,

1:19:39 in terms of,

1:19:40 um,

1:19:41 of the,

1:19:41 the effects that we find on the for the contribution rate and what,

1:19:45 what's driving them,

1:19:46 uh,

1:19:46 I think yes,

1:19:46 time preferences will be,

1:19:48 will be,

1:19:48 uh,

1:19:49 will be a factor obviously,

1:19:51 uh,

1:19:51 because that's going to reduce the how much value those pension contributions,

1:19:55 but also what I was trying to,

1:19:56 to,

1:19:56 to say which is that,

1:19:58 that,

1:19:58 um.

1:19:59 You know,

1:19:59 how far are people from the margin of indifference between formal and informal,

1:20:04 so to the,

1:20:04 you know,

1:20:04 to what extent are,

1:20:05 are people just um excluded from formal work because they don't

1:20:09 have the right skills or because there's not enough jobs or,

1:20:12 um,

1:20:12 or

1:20:13 what are the roles of these non-pecuniary preferences.

1:20:15 So all of,

1:20:15 all these are parameters in the model that,

1:20:17 that,

1:20:17 that play into

1:20:18 the,

1:20:19 the margin that we,

1:20:20 we ended,

1:20:20 end up measuring.

1:20:22 Um,

1:20:23 in,

1:20:23 um,

1:20:24 in terms of validating the,

1:20:25 validating the,

1:20:26 the results,

1:20:27 so typically,

1:20:27 so in those papers,

1:20:29 um,

1:20:29 we didn't incorporate,

1:20:31 uh,

1:20:31 exogenous variation that,

1:20:32 that would,

1:20:32 uh,

1:20:33 would validate that,

1:20:34 that,

1:20:34 uh,

1:20:35 that variation that those,

1:20:36 uh,

1:20:37 those numbers,

1:20:37 but,

1:20:38 uh,

1:20:38 we,

1:20:38 we drove,

1:20:39 drew from the literature.

1:20:40 So for example,

1:20:41 in the,

1:20:41 in the,

1:20:41 the reform,

1:20:42 the 2008 reform,

1:20:44 uh,

1:20:44 was,

1:20:45 um,

1:20:45 analyzed also with reduced form methodologies that.

1:20:49 That exploit how different birth cohorts are are exposed differently to,

1:20:53 to the change in rules

1:20:54 and,

1:20:55 and the magnitudes uh that we find are,

1:20:57 are similar

1:20:58 um but of course what we can do then with

1:21:00 the structural model is is simulate designs that did not,

1:21:03 um,

1:21:03 that did not,

1:21:04 uh,

1:21:05 uh,

1:21:06 that were not implemented.

1:21:08 Um,

1:21:10 And Sergio,

1:21:11 uh,

1:21:11 I'm sorry,

1:21:12 I'm,

1:21:12 I'm not sure I understood,

1:21:14 uh,

1:21:14 your

1:21:14 question very well.

1:21:15 So how much are government taking into account this,

1:21:18 the

1:21:19 evasion in terms of setting their,

1:21:20 their,

1:21:21 to what extent the decisions of pension are related to the decisions on,

1:21:25 on taxes or provision of,

1:21:28 of health insurance,

1:21:29 disability insurance,

1:21:30 etc.

1:21:31 To what extent this is isolated or is viewed as

1:21:33 a package or because I mean they are all related.

1:21:36 Yeah,

1:21:37 I mean,

1:21:37 theoretically,

1:21:38 as,

1:21:38 as I think they're,

1:21:39 they're,

1:21:39 they're,

1:21:40 they're,

1:21:41 they're related even though how much value your contribution to a pension

1:21:45 versus your contribution to uh health insurance is going to be different.

1:21:47 For example,

1:21:48 if you're very present biased,

1:21:49 you might

1:21:50 not value pension contributions,

1:21:51 but you might value the.

1:21:53 Uh,

1:21:53 contribution to health insurance,

1:21:55 um,

1:21:55 and,

1:21:56 and,

1:21:56 and

1:21:57 again taxes is the same,

1:21:58 right?

1:21:58 Like there's,

1:21:58 uh,

1:21:59 or

1:21:59 you know if you're completely

1:22:01 present biased and your pension contribution you're gonna value it,

1:22:04 uh,

1:22:05 as a tax,

1:22:06 exactly as a tax,

1:22:06 right?

1:22:07 So,

1:22:08 uh,

1:22:08 from the government side whether they take the,

1:22:10 the government takes,

1:22:11 uh,

1:22:11 the decision to pension in an isolated way relative to other,

1:22:15 other provisions for social safety net or,

1:22:17 or taxes.

1:22:20 Yeah,

1:22:21 yeah,

1:22:21 I'm not sure.

1:22:22 I,

1:22:22 I,

1:22:23 yeah,

1:22:23 we can,

1:22:23 we can talk

1:22:25 we can talk,

1:22:25 uh,

1:22:26 one on one,

1:22:28 and,

1:22:28 um,

1:22:28 Meliss,

1:22:29 thanks for,

1:22:30 for,

1:22:30 uh,

1:22:30 I mean it's always great to talk to you because you have all these,

1:22:32 uh,

1:22:33 this knowledge of,

1:22:33 of what's going on everywhere,

1:22:35 um,

1:22:35 and,

1:22:36 and,

1:22:36 um,

1:22:36 and yes,

1:22:37 um,

1:22:38 uh,

1:22:38 in terms of informal sector,

1:22:40 uh,

1:22:40 professional organizations,

1:22:41 um,

1:22:42 what I was,

1:22:42 I agree there's a lot of anecdotal evidence.

1:22:44 That governments are are interacting with these these professional

1:22:47 organizations in many ways and practitioners know about it but

1:22:50 what we don't have is like some systematic

1:22:52 uh way of documenting what how many let just if if only

1:22:56 what fraction of the informal sector is affiliated to an organization so

1:23:00 that's where I think research needs to happen to to to make it more systematic,

1:23:04 um.

1:23:06 And lastly,

1:23:07 um,

1:23:08 I'm not sure I have much to say about like the,

1:23:10 the gig economy,

1:23:11 so I agree that,

1:23:12 that,

1:23:12 that it,

1:23:12 it poses a lot of similar questions that,

1:23:14 um,

1:23:15 um,

1:23:16 to,

1:23:16 to,

1:23:17 to,

1:23:18 to what we've been discussing about,

1:23:20 uh,

1:23:20 informality,

1:23:21 um.

1:23:23 In,

1:23:23 in terms of,

1:23:24 uh,

1:23:24 if financing pensions with,

1:23:26 with uh let's say tax on,

1:23:28 on consumption,

1:23:28 I think,

1:23:29 uh,

1:23:29 Pierre,

1:23:30 uh,

1:23:30 Pierre Bachas in the research group has research that

1:23:32 looks at

1:23:33 what fraction of consumption is informal or is,

1:23:36 is,

1:23:36 is done,

1:23:37 uh,

1:23:37 in informal means and therefore would escape taxation.

1:23:40 And so I think that's a very important parameter to

1:23:42 take into account in order to understand who would be,

1:23:45 uh,

1:23:45 impacted by,

1:23:46 by,

1:23:46 uh,

1:23:47 taxes on consumption that would try to finance pensions,

1:23:50 um,

1:23:50 so yeah,

1:23:50 I think it's a very interesting,

1:23:51 uh,

1:23:52 dimension to look at.

1:23:54 Great.

1:23:54 So Miguel,

1:23:55 you,

1:23:55 you,

1:23:55 you have the floor for to close.

1:23:57 Thank you.

1:23:57 I have 4 points to make before,

1:23:59 before I turn back to you to close.

1:24:00 But,

1:24:01 uh,

1:24:01 the first is on Prianka,

1:24:03 what Prianka was saying,

1:24:03 which I,

1:24:04 I think it is extremely important.

1:24:06 Uh,

1:24:06 so yes,

1:24:06 there is a lot of,

1:24:07 there are a lot of analogies between social health insurance and

1:24:11 pension insurance,

1:24:13 except that

1:24:14 I would still

1:24:15 insist on a distinction between.

1:24:18 Retirement pensions and all other kinds

1:24:21 of pensions like disability survivorship.

1:24:23 What is the distinction?

1:24:25 The distinction is that in the case of survivorship or disability,

1:24:30 there is a huge analogy with health insurance.

1:24:33 This is really a part of social insurance.

1:24:35 Um,

1:24:37 and,

1:24:37 and we can talk about similarities.

1:24:39 Obviously

1:24:40 maybe myopia is smaller in health,

1:24:42 but,

1:24:43 but then there is this

1:24:44 big issue of creating the health package.

1:24:47 I see retirement pensions as they look like in the world slightly differently,

1:24:52 primarily because it stopped being an insurance.

1:24:54 You can think reasonably about insurance if

1:24:56 the probability of something occurring to you

1:24:59 is,

1:24:59 I don't know,

1:25:00 it was like 100 years ago receiving retirement pension.

1:25:05 The age of eligibility for retirement,

1:25:07 1 in 6 workers was getting it,

1:25:09 1 in 8.

1:25:10 Now we have a situation where 9.5 in 10 get it,

1:25:13 so it doesn't lend itself to insurance type thinking.

1:25:16 It lends itself to savings type thinking.

1:25:19 This allocation of consumption over the lifetime is fairly predictable.

1:25:23 That's why I am a big fan of defined contribution pension schemes,

1:25:27 but I'm talking retirement pensions,

1:25:28 totally different ball game with other sorts of pensions.

1:25:31 So I want to make this distinction which I think makes it different than health.

1:25:34 In terms of retirement pension,

1:25:36 which of course

1:25:37 we need to look at what income level,

1:25:39 uh,

1:25:40 be that,

1:25:41 that,

1:25:41 that becomes blurred when we talk about formalizing the informal sector,

1:25:45 but certainly we need to do work together

1:25:47 and I always

1:25:48 wanted us in the bank to do more work together in the regions and globally,

1:25:52 and it hasn't happened like,

1:25:53 you know,

1:25:54 uh,

1:25:54 from our GP folks from HNP from DC sitting together and doing things

1:25:59 that would be,

1:25:59 we,

1:25:59 I don't think we've done enough,

1:26:01 enough of that.

1:26:02 Second point,

1:26:03 um.

1:26:04 Uh,

1:26:04 second point,

1:26:05 uh,

1:26:05 on,

1:26:06 on,

1:26:06 uh,

1:26:07 taxation and social insurance again,

1:26:09 the situation that we haven't done enough and that's why it is so exciting

1:26:13 what is happening in luck now

1:26:15 because this is essentially looking at

1:26:18 tax and benefit reforms

1:26:20 so that to improve

1:26:22 redistribution and risk sharing in social protection schemes.

1:26:25 So in this sense this,

1:26:26 this,

1:26:26 this is precisely when this is going to be brought

1:26:29 together as we will be discussing this next week.

1:26:31 So I'm very excited.

1:26:32 About it,

1:26:35 two last points.

1:26:36 One is Gustavo talked about vesting period that may be a problem,

1:26:40 but Clement,

1:26:40 we discussed a few days that the lack of vesting period

1:26:43 may be a problem because I didn't know that in Chile,

1:26:46 even if you have a very small savings,

1:26:49 it's still being converted into some kind of annuity

1:26:51 which makes it a mockery because it's so low.

1:26:53 So it would be better if this possibility didn't exist,

1:26:56 meaning unless you accumulate something

1:26:59 which is

1:26:59 in part of the.

1:27:00 Resting period,

1:27:01 right,

1:27:01 unless you accumulate through that,

1:27:03 then you cannot get a pension.

1:27:05 The moment you

1:27:06 anyway it's for separate discussion what's what's better,

1:27:08 what's the right vesting period.

1:27:10 But I understand too long of a vesting period is a problem,

1:27:13 but the lack of a vesting period is another problem with different consequences.

1:27:17 Finally,

1:27:18 my final point is that

1:27:19 the way I got Gustavo's comment was,

1:27:24 was

1:27:24 maybe I am now putting my words in his mouth.

1:27:27 It's about observability.

1:27:28 Of transactions,

1:27:30 given that so many transactions are

1:27:32 observable transactions in terms of consumption,

1:27:35 because mobile money is used,

1:27:37 could we use it as a tax base for social

1:27:39 protection purposes by automatic default that $1 goes somewhere?

1:27:45 If you do a transaction

1:27:46 like I heard,

1:27:47 I haven't seen it,

1:27:48 that in Ethiopia,

1:27:49 not Ethiopia,

1:27:50 in Estonia,

1:27:51 in cafeterias in Estonia,

1:27:52 you have schemes when you save $1 automatically as a default if you buy coffee.

1:27:57 Because

1:27:57 it's observable and you use your,

1:27:59 your card,

1:27:59 of course you can opt out,

1:28:00 but the point is the default is there.

1:28:02 So and you can choose whether you want to be in this.

1:28:05 I don't know the details,

1:28:06 but I think that line of thinking,

1:28:09 uh,

1:28:09 in addition to thinking about what's

1:28:12 the right subsidy rate from general revenue to contributions,

1:28:15 that's one thing,

1:28:16 but the other thing it could we automatically be taxing transactions.

1:28:20 And then earmark or not to earmark is yet separate discussion,

1:28:23 but I think

1:28:24 we see it often in our discussions as

1:28:26 a discussion that has big potential for coverage of the informal sector.

1:28:30 And thank you very much again for having me here.

1:28:33 Really great birthday gift.

1:28:35 So much appreciated.

1:28:36 Thanks,

1:28:36 Michel.

1:28:37 Thanks,

1:28:37 Clema.

1:28:37 Thanks to all of you for great questions and,

1:28:39 and,

1:28:39 uh,

1:28:40 we'll see you next month for our next PRT.

1:28:41 Thanks a lot.

showAllTimestamps

no

transcript

Um, Welcome. Uh, my name is Dion Film. I'm the director of the World Bank's research department. So, uh, welcome to this October edition of our policy research talks. Um, as some of you may know, these, uh, provide us an opportunity to present work coming out of the department with the goal of sharing the findings inside and outside the department as well as with others outside of the World Bank. Uh, with that, I'd like to welcome our audience in the room today, uh, on the Webex as well as on YouTube. Uh, today my colleague, uh, Clement Joubert is going to, who is an economist in the human development team in the research department, he's going to discuss, uh, how his and other research informs pension policy design in high informality settings, and he's gonna cover issues such as the disincentives for formal work, gender disparities, and the extent to which workers in the informal economy can participate in voluntary pension schemes. Uh, it wouldn't be a research talk without a conclusion that points to further research needs, so I think he's going to conclude with some unanswered questions and knowledge gaps. Um, it's my great pleasure to welcome, uh, Mikhael Ruovsky as a discussion today, not least because he was one of the very first people I worked with when I joined the World Bank to work on the 1995 edition of the World Development Report entitled Workers in an Integrating World. Um, Michel is the global director for Social Protection and Jobs, overseeing the World Bank's work in developing systems that protect the poorest and vulnerable from crises and shocks, as well as supporting private sector-led growth. Until July 2016, he was the director for multilateral organizations, and prior to that he was the country director for the Russian Federation and the World Bank's resident representative in Moscow for three years. Before that, he held several positions including sector manager for social protection as well as country economist for the Russian Federation, in addition to the pinnacle of his career, which was being a core team member of the 1995 WDR. Uh, prior to, uh, working, uh, to joining the World Bank, and this is actually very relevant to today's, uh, discussion, he helped design the Polish pension system while serving as the director of the Office for Social Security Reform in that country. He also taught at the Institute of Economic Policy at the Warsaw School of Economics. So the session today, I'll ask Clement to talk for approximately 45 minutes, after which we'll hear from Michel for about 10 to 15 minutes, uh, to get his reflections. We'll conclude the session with Q&A from the audience. Um, if you have a question, please use the raised hand option in Webex or signal to me in the chat that you have a question, and I can call on you. If you're in the room, just raise your hand. Um, if you're on YouTube, please put a question in the YouTube chat. I won't see it, but it will get relayed to me. So, uh, I can, I can, uh, relay that to, to Clement or Michel. Um, just a reminder, we are recording the session, um. And also, if you're online, please mute if you're not speaking. So with that, let me turn it over to to Clement. All right, um, is the microphone working properly? No, hello. Oh, there you go. Um, thank you for the introduction, Dion. Uh, thank you, Michel, for agreeing to discuss this, this talk, and, uh, good morning, everybody here in the room and, um, and online. Um, so my name is Clement Joubert. I'm in the, in the research group, um, and today I'm gonna draw from three of my research papers. Uh, to discuss issues related to pension design in context where informality is prevalent. Um, You probably are aware that that most workers globally do not contribute to pensions. In fact, about 2/3 of workers worldwide do not contribute towards an old age pension. Um, and at the same time, uh, we know that low and middle income countries are aging, and they're aging faster than, uh, current developed countries did in the past. So we can expect in the near future that fewer active individuals will have to support a growing elderly population. Um, and that this will put a strain on traditional old age, uh, support arrangements. So, um, encouraging long term savings, uh, for example, through, through by extending pension coverage, uh, could be a way to, uh, finance the future costs that are associated with an aging population. Um, now, the picture is actually grimmer than those numbers, um, um, suggest because, uh, when you start looking at, um, at how people contribute over their life cycle, you find that there's even fewer, uh, consistent contributors. Uh, pension economists, uh, talk about contribution densities, uh, to describe the fraction of the months since, uh, the, uh, since a worker enters the labor force where, uh, they contributed. And um if we look at data from uh from Chile, so on the right you see uh the distribution of these contribution densities among men and women um in Chile, you can see, um, you can see bars at 0 and 1. Those are people who never contribute or who always contribute, but you also see a lot of people who, um, who switch in and out of contributing. Um, so a lot of switchers for women, you know, labor market exists, exits further reduced contribution densities, um, and so that means that if in the cross sections, or, so, for example, in Chile you have about 2/3 of workers who contribute in a given year, uh, once you start looking at people who contribute, uh, year, year in, year out, you're left with maybe 1/3 of, of, um, of, uh, individuals contributing. Um, another difficulty with pension designs in, in this kind of, uh, setting is that you have what you could call a vicious circle of contribution evasion. Um, so obviously pensions, um, ask you to pay a cost today in exchange for a benefit far in the future. And so as soon as you have, uh, any, any form of present bias or even, um, just liquidity constraints, uh, this is going to create incentives for workers to avoid those, uh, those payments. And once you start having contribution evasions, you're going to get a pension system that produces low pensions or maybe that is in deficit. And then what you need to do in that case is either you're going to have to increase the, the contribution rate to prop up pensions or you're going to have to raise money through through taxes and in both cases that can further uh uh um aggravate the evasion and that was very much the situation uh in Chile before the famous privatized pension system was, was implemented in 1981. Um, So in this talk I'm going to consider 3 main questions, um, drawing from from 33 research papers. So first I'm going to present some results on the extent to which pension pension system rules generate informality themselves. Um, second, I'm going to talk about, um, gender pension gaps and how the design of the pension system can, uh, can help reduce them. And lastly, I will talk about um whether informal workers in the absence of pensions are uh saving uh accumulating wealth or saving um or not. Um, so the first two, questions will, will be in the context of Chile, um, which is at the lower end of the spectrum in terms of informality rates, but, as we've seen they grapple with very, very much with these issues of pension coverage, um, because of these, uh, inconsistent, uh, contribution patterns, and the third, uh, paper will draw, uh, uh, we use data from Pakistan which is, uh, uh, uh, a setting where informality is nearly universal. Um, and, um, and, um, you know, so the, the second paper is, is, uh, co-authored with Petra Todd at the University of Pennsylvania, and the third paper is, is, uh, co-authored with, with, uh, Priyanka, uh, who's here in the room. So thanks for coming. Um, before I get into the results, I would like to, to, uh, introduce, um, a, a small framework to, to, to show how, how I think about, um, the choice between, uh, formal, formal and, uh, and informal jobs, uh, and that will help us, uh, understand the, the results later. Um, so if we consider a worker who is either choosing, choosing between a formal job and an informal job opportunity, or maybe, um, you know, he's trying to decide whether to register his business or not, um, we can write the value of working in, um, let's say in the formal sector, um, as, um, including three main components. Um, so the first one is the utility derived by, by the wages that you can, uh, obtain in either, uh, through other, either option, but there may be also a lot of non-wage considerations. Uh, you might think that some people argue that the informal sector gives you more flexibility. Uh, on the other hand, formal jobs might give you access to amenities or, or, uh, or to, to, um, let's say health insurance, so you might have these non-wage preferences for working for, uh, one sector versus the other, and these might different, uh, for different workers. And lastly, you're going to have dynamic considerations. So if you work formally today, that's going to have consequences for you in the future as well. Um, you might accumulate different sets of skills. Um, you're going to accumulate pension savings or pension rights, uh, if working formally. Um, it might be the case that if you work formally today, that's going to make it easier for you to keep working formally tomorrow, so that's going to increase your, your, your options in the future. And um you might, um, it might help you um achieve eligibility for some um government transfers. Now, if we zoom in on the the wage part of um of of this equation, um, so obviously if you work formally, your disposable wage is going to be reduced in proportion to the contribution, the pension contribution rate because that money is going to be uh either stored away in an individual account or is going to be collected by the government in exchange for future pensions. Um, and then, um, your earning opportunities in formal or, or informal jobs will depend on the type of skills that you have, both schooling, um, the experience you've accumulated in either sector, um, also you might have an innate ability to work, um, to be productive in, in, in, um, in formal or informal jobs, and then you also have, you know, um, time varying opportunities that, that arise, uh, in, in either sector. So all of this to say that the value of working in either sector will be heterogeneous in the population. Um, and so in when we think about the impact of pension rules on these decisions, what is going to matter is, is the relative value, evaluation of formal versus, uh, informal wage opportunities for each worker, um, in each point in time. And, um, so this is, uh, written as VF minus VI. Um, and you might have different, uh, you could imagine different cases, right? And on, on the left hand side, you have a situation where workers are relatively homogeneous and they are, um, close, relatively close to the indifference point between working formally or informally. Um, whereas in case two you have workers who are heterogeneous in how they value, uh, each sector, uh, with some workers who are, who value formality very much, maybe because they can earn a lot more in the formal sector or because they value health insurance a lot, um, and other workers who, um, value informality a lot, and that could be, for example, because they don't even have access to formal jobs, right? That could be a way to describe it. If we introduce, if we change the rules of, of the pension system in a way that makes formality a little bit more attractive, um, in case one, you're likely to displace a large number of workers into formality, whereas in case two, the margin is going to be much thinner and, and, um, you can expect your, your effects to be, uh, much lower. So it all depends on how these, these, um, um, productivities and, and preferences are distributed in the population. Um, and last little bit before I move into, uh, move on to results, um, the results, um, for the first two papers at least, um, are, um, are, uh, use a methodology called, um, a structural estimation, and without going into details, I just want to give you a sense of, of, of what it is so that you, you can interpret the results better. Um, what, uh, you do first is you develop a model. In this case it's a model of how, uh, individuals or households are saving and choosing and, and, and supplying labor over their life cycle. And the key, as we've just discussed, is to introduce enough heterogeneity into the model that you can capture this distribution of, of how people value formality versus informality. Um, then you estimate the parameters of the model using, um, panel data in this case of on Chilean households, and once you have this, this, this, this, uh, this tool, this estimated model, you can, um, because the rules of the pension system are explicitly modeled, you know, the, the pen, the pension system contribution rate is a parameter in the model, you can switch those pension rules or move these pension rules around and simulate what happens to labor supply saving decisions. Uh, but also pension benefit levels and government costs under different design. OK, so, um, 1st, 1st question that, uh, we're going to look like, look at is, um, how do pension rolls, uh, do pension rolls rather, um, contribute to, to informality? Um, So what we do here is, um, using this kind of estimated structural model, we're going to change the contribute, we're first going to change the contribution rate around. Um, and the reason we're interested in the contribution rate is because, um, you know, contributions, pension contributions are essentially mandatory and illiquid savings, and to the extent that, um, you're forcing people to save more than they would have, um, on their own, this is going to, um, create this utility and essentially reduce the value of working formally. Um, so on the right hand side you can see the, the effect of moving the contribution rate around from, uh, a value of 5%, which it would be a very low contribution rate, um, all the way to 20%, um, so that kind of spans, uh, the, the typical contribution rates you see around the world and the, um, Chilean, in the Chilean case, which is the, the source of these, these numbers, um, the, the contribution rate is at 10%. And what you can see is that in the left hand side of the graph, the first margin that is activated as you move the contribution rate around is the substitution between pensions and non-pension saving. So the first thing that that uh households can do is, uh, if you, if you ask them to contribute more towards pension they might reduce their savings elsewhere but uh at some point, uh, they're going to either run out of savings to substitute from or they are going to want to keep a liquid savings so that they can. Um, self-insure against, against short term shocks and so that margin kind of shuts down and what you start seeing is more people switching to, um, to informality and, and, or households just producing their current consumption. So overall, uh, the impact that we find of increasing, uh, the contribution rate by 10% points to give you an order of magnitude is, uh, an increase in informality by about 8.3% points for men and 5.3% points for women, um, so. You know, essentially what that means is that this is, these are significant or sizable numbers, but at the same time it also says that, you know, the informality we see in Chile is not all generated, they're generated by the, by, by, by the, the, the obligation to contribute. Um, one interesting bit as well is that these effects are concentrated, uh, among workers with lower contribution density. So, uh, essentially workers who always contribute, uh, that the fraction of workers who always contribute does not change, um, and really it's those occasional contributors who are, are pushed out of, uh, formality. Um, a second feature of the pension system that has been, um, accused or, or, or, or, um, thought to, um, generate informality is minimum pensions. Um, and so the mechanism here is that minimum pensions implicitly tax contributions. So this is, uh, easy to see when you, when you consider a very simple minimum pension. Um, so, you know, the, the black line on this graph is your contributory pension. So the more you contribute, the higher your, your pension, and let's say the government wants to guarantee that pensions are not going to drop below a certain minimum. Um, in that case, um, for people who expect to qualify for that minimum pension, um, it's, it's quite obvious that contributing one extra year is not gonna raise their pension level, and in that sense, those contributions, those pension contributions are taxed, uh, implicitly at a 100%, uh, rate. Um, so that's going to create disincentives for, uh, this segment of workers to, um, to work formally. Um, and so it's been proposed and in fact implemented in, in, in some reforms, uh, to, to try to reduce these implicit tax rates, um, and so the way, uh, you can do that is by, um, tapering the minimum pension benefit and what we mean by that is, is moving from this, uh, red dashed line to the solid line. So what, uh, what you're doing here is, um. As people contribute more, you're reducing the government transfers, so you're giving them a little bit less, but it's not a one for one substitution. So you, as they contribute more, they're still increasing their, their pension a little bit, but they're still implicitly taxed, uh, at a certain rate. Um, um, and so, uh, we're going to anchor our analysis again on, on the Chilean case where in 2008, um, the minimum pension was reformed and they adopted, um, a design where the implicit tax rate is about 30%. So, and uh the policy experiment that we do specifically is that we're going to fix um the level of um the intercept of that of that graph so the level of the basic pension and we're going to move the slope around so vary the implicit tax rate and see what happens to to informality and what we find is that actually um the impact is very small so informality doesn't change that much. So this, which is kind of contrary to the, the uh ex ante expectation. So what is happening here, um, it's true that you are lowering disincentives for people who would have, um, um, qualified for the old minimum pension. Um, but at the same time, now you're going to have more recipients, right? It's pretty obvious from the, from the graph as you move from the dashed, uh, red line to the solid line, you're increasing the range of, of, uh, of, uh, of, or the, the, the, the set of people. Who are going to qualify for a minimum pension. So essentially what you're doing is you're spreading this implicit tax rate over more people. So you're reducing the tax rate, but you're, you're, uh, incorporating more and more people into these incentives. And so even theoretically the effect is, um, is, uh, is not, uh, is, is ambiguous. On top of that, you might expect that um contrary to what's happening with the contribution rate that we talked about before, this is um this is very far in the future for many workers, right? So if you're 20 years old and you're trying to decide between formal and informal, you, there's still a lot of uncertainty uh about your career that uh prevents you from knowing exactly if you are going to qualify for a minimum pension or not, uh, even, even in the absence of any, um, present biased or anything like that. Uh, so this uncertainty is going to dilute these disincentive, uh, especially at younger ages, and the last little bit that explains this result is that, um, you are transferring more money, uh, to the, to, to from, from the government, as you can see, you know, the, the, the government transfers can be loosely captured by the shaded area below the minimum pension line and above the contributory pension, so you can see that as you move, as you, as you, um. You make the, the, you increase the slope of the minimum pension, you're going to transfer more money to the population that's going to create wealth effects and disincentivize um uh formal work. All right, so just to summarize this, uh, the, the, the set of results, um, um, if we look at the contribution rate, we do see, uh, we do predict a sizable effect, uh, on informality, an effect that is likely larger for younger workers who are not yet locked in either the formal or the informal sector. Uh, note that obviously informality is not the only and perhaps not even the main consideration in deciding what your contribution rate should be. You know, pension adequacy would be the first, uh, thing that comes to mind, but also maybe more subtly, um, if you lower the contribution rate, maybe you're going to reduce informality a little bit, but you also are going to increase the cost of your safety net because people are going to accumulate less money or less pension rights into the system, and so they're more likely to qualify for whatever minimum pension you have there. Um, on the other hand, uh, minimum pensions and particularly the, the, the implicit tax rate or the way you're designing them, um, is likely to have a small effect on, on informality, um, and if to the extent that they exist, they are going to be close to, to the retirement age when people know exactly, um, or know much more about their, their eligibility. Um, other considerations, as I said, you're transferring a lot more money, uh, to, to people under a more tapered design, and so that's going to have a first order impact on the cost of your, your minimum pension for a given, uh, you know, consumption floor that you're trying to guarantee. All right, so 2nd, uh, 2nd set of results, now I want to talk about, uh, gender pension gaps. And this is drawing from a, from a different paper but still in the context of, uh, of Chile. Um, so gender pension gaps are pervasive. They, they are, they exist in most, uh, pension systems around the world, and the reason is because there are, um, several compounding factors that, that contribute to them. Um, the first one is labor market opportunities. Women tend to be paid, uh, lower wages. They also, um, may have, uh, lower access to formal job opportunities. Um, also, um, women tend to exit the labor force when they have children, and that's going to reduce their, um, their contribution densities. But also, um, uh, there are rules in pension systems that are, uh, not gender neutral. Um, so many, not all, but many, uh, systems have different retirement ages for women versus men. Um, and particularly in DC schemes, um, uh, the, the, the, the pensions might be calculating using gender specific mortality tables, and because women, uh, live longer, that will result in lower, uh, pension benefits for them. Um, so we're going to examine two strategies for tackling the gender, uh, pension gap. And the first one is to just expand the, the minimum pension, um, because women are more likely to have few years of contribution and low pensions they are going to to benefit a lot more from this type of, of policies and in fact that's what happened in uh in Chile in 2008. The, the system was reformed. Um, targeting, uh, specifically gender pension gaps, um, and, um, the way they, they, they, um, they try to achieve it is by increasing the minimum pension level, removing contribution requirements, and, um, also widening the means test that, uh, um, that you need to pass to, to be eligible. Um, and, uh, this was, uh, this, um, uh, effectively reduces the gender pension gap, um, by, by, um, by a large amount, almost, uh, um, uh, splits, splits it in half. So, um, going from 44%, uh, gap to 23% gap. And we also find that the effect is largest for women with lower wages and lower attachment to to formal work. So, um, women with lower schooling, uh, single women will um will see the biggest reduction in the gender pay gap. Um, the second strategy that we, uh, simulate, and that one was not, didn't make it into that reform, is to, um, to equate retirement ages. Um, so, uh, currently women in Chile can claim a, a pension at 60, um, instead of 65 for men, and we look at what happens if we equate those two, those two numbers, and we find that the reduction in the gender pay gap is also very large, uh, in fact, of the same order of magnitude than what, uh, we found, uh, um, with the minimum pension, uh, um, policy experiment. Uh, but now, uh, it's a very different, uh, groups of women are affected. So now, uh, you're reducing the gap among women with higher schooling, um, and, and, and, uh, and single women. Um, and the mechanisms that explain this impact is that when you, uh, uh, increase the retirement age, first you're reducing longevity at retirement. So, um, in, in systems like Chile's, you, you have a, you accumulate a certain certain amount of money and then you buy an annuity or something, uh, similar to that. So the longer you're expected to live, the, the, the smaller your, your, um, your, your benefit. Uh, in addition, you might expect that between 60 and 65, some, some women are gonna work and work formally, so they're going to contribute to their, to their account, but that's actually not the main, uh, the main, uh, explanation. And then lastly, uh, pension savings are left for 5 more years, uh, to accrue interest, and that's also going to increase, uh, increase pensions, um. Right, so, um, just to summarize, uh, summarize these, uh, these results, uh, we find that minimum pensions can be a powerful tool to reduce gender inequality, uh, gender, gender pension gaps, uh, with an effect that is concentrated on, um, on low schooling women and homemakers. Um, so one thing that we look also a little bit in the paper is that, um, you know, when you think about distributional effects of, of this type of reform or even disincentive, uh, uh, uh, effects, um, you have to take into account the household structure and the fact that. Um, the women with the least pension rights in Chile and in many countries are homemakers who, who just never contribute, but they might be married with middle class or even high earning husbands. So in terms of, of, uh, distributional effects and, and disincentives, you, you need to, to take that into account. Um, and we also find that the retirement age has a first order effect on the gender pension gap, um, but that is concentrated on high schooling and, and single women, um, and here also you kind of wanna be careful because, um, essentially if, if you're, um, if you're considering, uh, a rational agent, right, what you're doing is you're taking options away from that person, right? You're forcing them, uh, to, to retire you're, you're not allowing them to retire early. And so that's not going to be a welfare enhancing for women and in fact that might be part of why this reform hasn't made it to to uh or this feature hasn't made it to the reforms, uh, even though it has been considered. Um OK. And so, uh, now I would like to, to, to move on to the 3rd set of results and the 3rd paper. Um, so, um, as I mentioned, this is joint work with, with Priyanka Kant, who's, uh, who's working on health in the MENA region, uh, now. And uh we're moving to the context of Pakistan and what we want to talk about is um um is about informal sector pensions. Um, but you know, backtracking a little bit, um, you know, if we look at, at, um, uh, at, uh, many developing countries, um, or let me start actually with, with, uh, you know, the recent, um, social protection and jobs compass which uh sets as a, as a long term goal universal social protection. But in many countries what's going on at the moment is that you have the poor who are targeted by social assistance and uh formal workers who are covered by social insurance programs and then uh in the middle you have the non-poor informal segments um so in a in a in a report that I co-authored with with Medis who uh Medis G who's here in Himani uh Jain. We take data from 6 countries in Africa and we, we map out the segment and, and look a little bit at at their characteristics and, um, you know, in, in these, in these countries, um, the non-poor informal segment is more than 50% of, of the population. And we also find that it spans, uh, uh, quintiles, uh, including the rich richest quintiles, which is interesting. Um, we can also anticipate, um, uh, as poverty, uh, rates decline and formality doesn't look, uh, to decline or is stagnating, that the segment might, uh, increase, uh, over time. Um, so, um, a growing number of countries are trying to, um, target this segment, uh, by introducing voluntary pension schemes that are, are designed specifically for, for the informal sector. However, despite a few, uh, some promising, um, promising cases, um, most of these schemes have struggled to attract participants. Um, so we can come up with a lot of reasons for why that might be the case. Um, it might be that, uh, workers are not aware enough of these schemes, uh, that they don't understand them because of financial literacy. It could be that they don't trust the scheme and don't want to store their money in there. Uh, there might be a lot of behavioral obstacles to, uh, to registration and participation, um, but the explanation that, um, we wanna start, um, looking at here is maybe something more fundamental is maybe workers, uh, or informal workers just don't have the ability to save and particularly to save for the long term. Right, it could be that they just are not, uh, they have low disposable incomes, or they are, they have uh their exposure to risks and their liquidity needs are too high to um afford to, to store money in a, in a pension account. So, um, this ability to save is, is something, is a concept that is very hard to, to get to. So perhaps the first step, uh, um, is, uh, what we're doing in this paper, which is to just measure what wealth accumulation looks like in the absence of a pension scheme. And that will give you a first indication of the ability and desire of, of households to set, uh, money aside for the long term. Um, to do that, we go to Pakistan and the reason is because, um, um, there's a household survey in Pakistan that asks consistent, uh, wealth questions, uh, over more than over 18 years. Um, and let me tell you, I mean, it's a well known problem, but let me tell you why you need that much data. So if you look at wealth accumulation or in a given year by comparing younger, uh, household with a younger head and households with a, with an older head, um, you find some, a profile that is relatively flat. So that's the dashed line, um, in the graph on the right. Um, so this, this is, uh, household net worth, um, in 2010, right, um, by the age of the household head. Um, if you have data from many surveys over time, you can now track cohorts, and, uh, you can see, for example, uh, here in the, with the 1960 birth cohort, that the profile suddenly looks a lot steeper. Um, and so when we actually, um, decompose, uh, uh, and, and isolate the, the effect of, of, uh, age of the household head on net worth, we find that between the ages of 625 and 65, a household accumulates the equivalent of 4 years' worth of consumption. So the idea being that with that, uh, uh, wealth accumulation, if they sell everything, they can survive for 4 years, just to give you the, the, the thought experiment here. And then we, um, we dig a little bit more into what type of wealth is accumulated and find that, um, uh, almost all of that accumulation happens in housing, so 80% of it, and by contracts, for example, land is a sizable fraction of the portfolio, but it doesn't move over the, uh, with age that much. And other types of uh um instruments that we measure, uh, durables, uh, businesses, uh, financial wealth, loans that um changes, but the amounts are, are dwarfed by, by housing. Um, we also look at, um, accumulation rates, uh, over the consumption distribution. And um and what you see here is that uh the the rate of accumulation is actually very similar across consumption quintiles. uh you see that the first quintile accumulates a little bit slower than the others, but other than that, um, the rates are similar. Now, um, of course this is relative to household consumption, right? So clearly, uh, wealthier households are going to accumulate. Um, more wealth, but relative to, uh, their level of consumption, the, the accumulation looks quite similar. OK, so, uh, to, to conclude on this, um, you know, this is a descriptive paper, uh, we just wanted, wanted to put a number on, on this question of, you know, of wealth accumulation when there's no pension system, and we find that, uh, accumulation is sizable, uh, about 4.2 years' worth of household consumption, uh, over, over 40 years. Uh, we also find that that the rate of accumulation, um, gets faster after, after age 40. We find that most of the accumulation involves housing wealth, which is interesting because it kind of gives you also uh a sense of the need for liquidity, right? Like we, we, it indicates that households are able and willing to save in a relatively illiquid form so housing is less illiquid than let's say a pension account, but um it's still a relatively illiquid asset. Um, and we find that accumulation, um, looks, um, looks similar in relative terms between the poor and the rich, which may be important because, uh, the question is, do you want to target, let's say those informal pension schemes only to the wealthier parts of the informal sector or to all of the informal sector, uh, so this can, can, uh, can, um, speak to, to, to that, uh, that problem as well. All right, so this gives me, I think I still have a bit of time to, to, to give you, um, uh some final remarks on, on things that I think we don't know as much about and that, um, I hope to, to, to be thinking about in the future. Um, so going back to this non-poor informal segment, um, you know, there's been a lot of research since the 90s on the poor and on their financial and economic lives and we've learned a lot about, about that. There's also some research on the near poor and on transitions uh from, uh, from being not poor to being poor, but. Um, there's still a lot of things that we don't really know very well about, uh, the, the non-poor informal workers, and this is going to be key as we try to move towards universal social protection because these are the people that we still are left to, to be covered. Um, so for example things like transitions between formality and informality over the life cycle are not, uh, typically well described in most countries. Um, we tend to think of the formal and informal sector in a static way as, you know, you have the formal ones and the informal ones, but to what extent are, are there transitions, um, in Chile as we see there's a lot of switchers, um, so that's going to matter for pension design. Even things like shock exposure, need for liquidity, ability to self-insure, um, are often described, well described in the cross section, but, uh, we don't know very well to what extent it's the same households who are, uh, chronically exposed to shocks or, uh, whether, uh, it's, it's uh more widespread. Um, and then expanding on the results on, on housing wealth, um, you know, the question now becomes, uh, do households save in housing because it's optimal for them for whatever reason or because it's a lack of option. And, and the, and therefore, you know, would it be beneficial to, um, uh, fix that labor that uh financial market failure and, and give them another instrument where they can, um, put some of that, uh, of that wealth and would, would they gain gain from that. Um, and another, um, another topic that, um, I've, I've started looking at with, uh, with, uh, Kathleen Beagle at the, at the research in the research group is the question of professional organizations in the informal sector. So we tend to think of the informal sector as, um, atomistic, uh, you know. Uh, um, individual workers, but in fact anecdotal evidence, uh, and, and you know, many, uh, practitioners that we talked to during our cross support, uh, have a lot of things to say about, uh, informal sector professional organizations. So, um, the informal sector is organized. Uh, but we don't know much, uh, at least in terms of systematic documented, um, hard evidence, uh, in terms of what, what they do, these professional organizations, and to what extent they might be providing some insurance that, and we, we would want to take that into account for. Um, for, um, in, in, in, in, in, in our operations, um, and also to what extent can they be leveraged to, uh, to help deliver social protection programs, uh, to the informal sector, so they can help identify beneficiaries, uh, communicate to beneficiaries, uh, help build trust in the schemes, uh, maybe even more directly collect contributions, um, so I think that there's a lot of things that we, we, we need to learn about, about this. Um, and just to, to finish, um, another kind of set of, of questions that, that, um, that I'm interested in, um, so as we know, population aging is, is accelerating and, um, we kind of know theoretically what needs to happen in order for, uh, countries to prepare for that, right? So we, we need a lot of investment so that the, the, the, we can finance, uh, the cost of, of, uh, of aging. Um, but I think what, what we don't see as much about is to what extent, um, decisions, uh, household decisions are adjusting to population aging, um, at both their decision and in fact their perceptions, to what extent households are aware that they're living in, in an aging world and anticipating that, um, you know, our individuals, um, expecting to work and live longer than, um, than their parents. Um, are they expecting to receive lower transfers from relatives, uh, and networks in old age? And are they preparing for that, um, as a result? Are they saving more or are they saving differently than, than, um, their parents did? Um, so for example, in Pakistan, even though we focus on the, the age dimension, we also see that. Um, saving also, um, uh, increases across cohorts. So the more, uh, recent cohorts to reach 65 years old tend to, to reach that age with, um, with more, uh, wealth than, than the, the previous ones, um, and same thing our households investing more in children in, um, because they anticipate, uh, that they're going to need their children to, to be wealthier in order to, to support them. Um, and lastly, our traditional care and co-residence arrangements, same thing, responding to population aging, uh, in any way? All right, so that's, um, that's all I have for today. I just want to thank you again and thank you, Michel, and uh don't forget to wish uh Michel a happy birthday on your way out. Thank you, Clement for that. Before turning over to Mihal, let me just remind you, if you're watching online, please just raise your hand or put a note in the chat that you have a question and I, I'll call on you, uh, during the Q&A session. So with that, over to, to you, Mihal. um Uh hello. Uh good morning, uh, everybody, and good afternoon to those who joined online from, from Eastern Hemisphere. It's a great pleasure to be here. It's something which people often say at the beginning, and I say it, but I have a reason to say that. I wanted to report to you that I'm actually quite often invited to some. Speaking engagements, but very seldom recently invited to such an exciting speaking engagement, which is commenting on a research paper. So I really value it even in annual meetings and all the retreats, it will stick out as a memorable experience for me compared to all others. So thank you very much for that. It's really a treat. Uh, let me say a few words about, uh, first about the context as I see it and then about the paper, uh, and while talking about the paper, I may, uh, mix it with suggestions for further research and, and policy implications. So the context is that the topic is, uh, is tremendously important and, uh, more important, uh, more important, uh, every year. Uh, and that happens because of, uh, persistence of informality in the world. Uh, in, uh, in our white paper a few years ago of who main author is in the room, uh, we documented persistence of informality and we also, uh, pointed out to the fact that in countries that became very formal, the traditional employment relationships are in decline, meaning, uh, working relationship become more fluid. Uh, there, there are many more transitions in the labor market. Uh, in other words, they start resembling more the informal labor markets which we know so well from, uh, low and low and middle income countries. So we have a persistence of informality in, in the low and low, uh, low and middle income countries and growing fluidity in, uh, in, um, in high income countries and, uh. In that context, covering informal sector by social protection has become a fundamentally important thing for us in the context which Clement described as, uh, our aim of universal social protection, which is an international aim, uh, we have the universal social protection 2030 movement of, uh, of which we are a part of. And um and the question is tremendously difficult uh we are thinking about the whole of social protection, but uh Clement focused on pensions here and pensions are probably the most important of all social protection schemes for the following reason that. As countries become richer, we see what's sometimes called SEC, social protection angle curve, which is that they spend more on social protection. As countries become richer as a percentage of their spending and percentage of GDP and then within the package of what they spent on social protection, they spend more on pensions than on other social protection programs. Again, comparatively speaking, so these are social protection angle curves here. So informality and pensions is important it is, it is already today, it is a topic of very fast growing importance. So, um, um, and then you heard from Clement about this group of non-poor informal which indeed remains, uh, has, uh, is, is, is of particular concern because when you build social safety nets, last resort social. Assistance programs, they typically concentrated on rural poor and when you think of existing formal schemes, they focus on those who are formal. So non-poor informal are neither here nor there, and that's why it is a group of concern. Moving to uh to the paper itself, um, I, and the, the, the let me talk in maybe 44 parts about 44 findings as I see it, so. So the first part about the contribution rates and informality, I really like the framework of, of Clement that informal sector is not really merely a residual labor market, but it offers competitive earnings opportunities. uh, and then in that context increase in contribution rates, uh, indeed increases informality and therefore reduces tax. Revenue, uh, I think that, um, especially with those with lower contribution densities, I, I, I think a method used is quite appropriate because we have this method of simulated moments which, as I understand it, is if you have really voluminous data from various sources and then you had, you had the longitudinal plus administrative, then this is a fantastic way of doing some kind of indirect inference and And uh when it comes to uh findings, which in a way would be expected, although they are quite strong here, um I think there are certain questions that come to mind for both further researcher in terms of policy implications. And the key question which is in my mind is as follows. So if Uh, can, can, can, let me also add something which I should have added up front. Any sort of coverage of the by social protection of the informal sector would probably have to take a form of a combination of contributory and non-contributory. A contributory being exactly what you examine Clement, non-contributory being general revenue, social assistance, subsidization of contributions or whatever form works. So we are talking about the world when this combination be uh needed. But the key question which needs to be added first, how much of that can be done by contributions, right? And the question that comes to my mind reading your paper is if countries were to take action to identify right contribution rates, not too high and not too low. So that they would have a desired effect. What would they be and how would they relate to both voluntary savings and optimal savings and voluntary and optimal are not the same as we discussed a few, a few years, a few days ago because of myopia, workers may not workers, informal workers may under, uh, under uh under save, uh, um, so, so, but the relationship between. Voluntary savings, optimal savings, and the contribution rates which are going to be used for contributory coverage of the informal sector is phenomenally important, important question to answer, uh, trying to answer a bigger question. Will the countries be able to increase coverage by contributory schemes in the informal sector without formalization and without being able to impose, to impose a mandate in in a in a hard way. Um, I like the, the, the second part of, of the paper and the issue of minimum pension, uh, and the Clobex story. Um, I think it is, uh, actually the fact that you identified 60, I call it benefit withdrawal rate, uh, what you call loback, but you identified like 60% being the best compared to other, which I, I reject, say forget about 100 or 0, right? So we are talking where are those ranges, and, and you found 60% is quite intuitive, um, uh, and I think the importance of um. Uh, of, uh, that finding, uh, is, is, uh, the importance of that direction of, of research is, is tremendous because, um, because if we need to cover the informal sector by the combination again we will be in the way to what are the, what's the right contribution. Uh, rate, but then because we are talking about the whole of social protection, all those who contribute are in the way in which they also, some of them receive social assistance, not contributory. But then you want to withdraw social assistance if the level of income goes up, right? So the question is what should be the benefit withdrawal rate to retain incentive to, to first of all, to, to, to work and then to contribute, and this benefit withdrawal rate or the clawback rate, it's not a clawback if it's maybe in the context of social assistance, but I think this is a hugely important topic and in fact both your findings, I think we are highly relevant in. The context of the work done in luck where I think Truman, you are leading it and I know Clement, you are in the team. I saw the team we are discussing it next Tuesday where the whole debate is about um it's called formalization but I call it uh coverage even without formalization how to expand coverage to cover informal sector which could be called that if they are covered and they are formal, right? But there are some fascinating parts. Of that analysis. One of them is the work on the formalization tax rate, which is the proportion of the change in earnings and which a worker loses when it moves into a formality which will be taxed away from, uh from him or her. I think that, that your work on, on, on this has enormous relevance for that part of the work on the, on the lab study. So I think there is a lot of uh potential, potential there. Uh. Your gender gap findings are actually the cleanest of all, because to me they are both intuitive. They are very solid, and in so many policy dialogues we are making exactly those points, maybe more about equalization of retirement age and how women would benefit from equalization, and then minimum wage point very well found in Chile. So is often made. So here we are in a situation where I think it's a little disagreement among researchers, among those who work on policies. However, there are big political obstacles. In fact, in fact, uh, speaking strictly, uh, the retirement age should probably depend on life expectancy, uh, uh, at the age of retirement, which would mean it should be higher for women. Uh, but I guess that that territory has not been entered in the discussion, but maybe it should as a starting point to broader, uh, to broader, uh, uh, to broaden the scope of discussion to arrive at reasonable discussion, but I really like those, uh, findings. And then when it comes to Euro-Pakistan analysis, which is also something which I think is, is, is, is, is extremely interesting. Uh, it does, uh, and I, I will not repeat the findings, but it poses, it, it poses actually a very interesting question which, uh, I, I myself thought very little about it, which is about that you have households who don't want to participate, but they do have sizable savings primarily in the form of, uh, savings in housing. And there's a lot of questions why it happens, what is special about that type of asset, but there is also a question about Uh, about, is there a way, um, is there a way to, uh, To use those savings in housing for the purpose of um for the purpose of uh retirement or disability and uh is there a role for public policy in doing here anything and I think this is in my vastly under researched and it is a topic of uh of big importance because we always always. In policymaking, we need to look at the whole picture, not only at pensions. We need to look at the whole of social protection, but then we need to add a whole of savings. So it's always, it's like, yeah, like an onion when you have those, uh, those, uh, those, uh, layers. And in fact, this, this, your, your, your last set of findings uh related to Pakistan it brings me to, to, to, to, to pointing out to the fact that the work on it, on the informality in the context of social. Protection we already talked about how important it is, but let me offer a few strands of what is actually happening. Medicine in the room is very much leading big strands of that of that work, which is that if. I already said first let's think about what could be covered by contributory and then obviously it's a question maybe we should subsidize it and where should be social assistance but starting with contributory, the big issue is what is the role of. Behavioral in others, what is the role of defaults to be set properly? What is the role of bundling services like you bundle. Old age security with disability security and health insurance together which may be more that can may make informal sector workers more willing to contribute if this is uh bundled and and how you allow for short term access along with long term access as Retirement pension savings need to be probably for that group accessible in the case of other life changing events adversely affecting income. So there is a huge set of, I think, research that needs to be done in, in, in this context. What, what is being done in our global practice, and I already mentioned is here that we are carrying surveys in several countries to. What you came out pointed out how it's difficult to understand, which is so-called willingness to save our ability to save, um, so, so that we hope that from those surveys we'll be able to figure out the willingness to save and that would contribute to our knowledge base which would ultimately may inform thinking about how to cover informal sector worker with contributory schemes or in what part to cover uh cover them um and then based on that what. Subsidization covered by general revenue would be appropriate and what what rate would be appropriate and this is something this is considered by some countries that combination and I think this is where the sort of nerve of further research should be. Um, and it should be there again, I repeat, because we cannot at all rely on formalization. It's, it ain't going to happen. It hasn't happened for 40 years. I love saying India was 90% informal in 1948. It's 90% informal now. Nothing changed there and trying to think in terms of formalizing would be like the definition of insanity, doing the same again, hoping for a different result, right? So you need a different approach. The approach we are using is let's cover them. Without formalizing or if in the process they formalize this even better, but the formalization should not be a center of a central point of reference because that may immediately lead us to wrong policy conclusions just by the way that we call it formalization. Let's let's tax them. Let's make it mandatory. Let's impose the law, which, you know, you can do all kinds of wrong things the moment you start, the moment you think how, how using behavioral knowledges, using soft mechanisms, encourage them becoming more formal with respect to social security contributions, because yet another issue is. Definitions of informality you can be formal with respect to taxes but not social security contributions or the other way around or formal with real estate taxes but not other taxes. I think there is a lot of work I've seen part of that in showing those different shades of gray. Let me conclude by Yeah, uh, and one thing which you said, which, which I think I seldom hear and I liked what I heard is you pointed out the role of professional organizations in the informal sector, and I seldom hear discussions on that. Maybe I, I just don't read the right things, but obviously intuitively speaking, I could think of organizations like SEWA in India and several other self-group. Of women, which played an enormously important role in the development of labor force participation there and could potentially play a very important role in expanding coverage of contributory schemes to the formal sector. So this is something which I'm happy you and Kalin are going to look at because it's tremendously important. So congratulations and thank you very much. The right discussions at the right time and the importance of that will be growing every year. Thank you. Thanks Michel for that. That's uh both the reflections also the, the very clear message about the role and the informality is here to stay, and we need to think about sort of policies in that context, um, before turning it back to, to, to Clement, uh, we have one hand raised, so I'm just gonna ask Gustavo, oh yeah, Gustavo, are you there? Um, and then I'll turn it back to Clement to res react maybe to, to, to Mikhael's point and maybe Gustavo's point, uh, question as well. So go ahead, Gustavo. Uh, thank you very much, Ion, and thank you, uh, Clement, for an excellent, uh, presentation but also for an excellent collaboration with, uh, with, uh, our teams on several, uh, um, fronts of, uh, the work on informality. I think this is, uh, this is, um, uh, a really an excellent summary of the analytical work, uh. Uh, and, and, and the link with operational work that, that, uh, you have been doing. Let me, let me, uh, make a couple of, uh, comments for your reactions. So you mentioned in the first part of your presentation, um. The uh the, the possible er um implications of uh regulations on uh on informality. And you focus mostly on uh contribution rates uh when when you mentioned other um um regulations such as minimum pensions you highlighted the importance to reduce inequalities but there are some uh also some concerns that minimum pensions and the. Expansion of uh social pensions can be a disincentive to formalize and I would like to get your your reactions on on on this. There's one point you didn't mention. You talked about the uh the contribution density and you didn't mention the best thing, uh, that, that the, the, the possible. This incentive of um investing periods of contributions so we have for example in Latin America countries where you need to have 10 years of contribution to get a pension and you have countries where you need to have 30 years of contributions to get a pension. Do you think these may potentially have. Anything uh to do with um the disincentives and to formalize when people think it's going to be unrealistic or not very easy to um. Uh, accumulate 30 years of contribution, but at the same time, uh, they may have easier access to minimum pensions or universal, uh, basic pensions and, uh, my second question is, uh, on your important contribution to understanding the savings potential of the informal sector and I think this is, uh, this is really a very fundamental piece and something that, uh, that has changed, uh, our. Uh, vision of uh of uh what solutions may work for the informal sector. So there's a lot of evidence and you accumulated more evidence you produce more evidence that that there is potential of savings in the informal sector. So, um, the question is, what else do we need to do to try to, uh, use that potential or to drive that potential or to encourage that potential to long term savings versus other forms of savings. Thank you. Thanks, Gustavo, um. Maybe Clement, if you want to react to sort of that collection of of of reactions and take it from there. Yes, um, thank you, um, both, uh, Gustavo and, and Michel for these comments. Um. So maybe taking the, the comments of Mijal first, I think, uh, uh, a broad, um, a broad thought that I have on, you know, choosing the optimal contribution rate and the optimal clawback rate and that's very obvious when you look at Chile is that, um, political economy considerations, uh, often trump whatever, you know, optimal rate you might find by doing the type of exercise that, that, uh, that I've done. Um, as you can see, you know, in Chile, the, it's not even clear that the system will, will survive, uh, in the long term. So, um, that's why I tried to, to mention a little, a little bit in the talk, but. Um, yeah, when you, when you, when you choose the optimal contribution rate, it might be that the, the, the, the optimal rate that minimizes disincentives to formality is X. So in fact, I, I think, uh, uh, uh, we do an exercise like that in, in, in, uh, I do an exercise like that in the paper, but, um, it seems, or one thing that, uh, that I've became convinced on, uh, of, um, uh, more recently is that. Um, you kind of need a contribution rate, or it has a value of also of a symbol, and so you need a contribution rate that leads to a pension that people can live on and, you know, even though, you know, in our, in our model, we could, we would assume that people can, can complement their, their, their, um, lower, a lower contribution rate with uh voluntary savings, but I think the, the, the contribution rate of the, of the system kind of needs to lead to. To a, a, a pension that people can, can, can live on. So I think, and, um, um, and after that you can, you kind of wanna take into account that it might create some this some informality, etc. but, but I think that's a, a primary consideration and likewise for the, for the clawback rate kind of a similar point is that um. You know, in the paper we kind of see, you know, the exercise that we do is to say if the goal is to, to, uh, sustain, uh, um, uh, consumption floor, what's the cheapest way to do it in terms of choosing that clawback rate, and, and we find that you want a relatively aggressive clawback of 60% or even higher, um, but, uh, and, and in Chile they, they chose a much lower clawback rate. Um, and, but part of that could be again political considerations, right, because you want to transfer some money to the, to the middle class and you do that by having a relatively soft clawback rate so that, uh, even people with relatively higher pensions are gonna get something from the government, um, so again, I mean also in this case really uh political economy considerations are going to be, to be very important to take into account. Um, Let me see. So, so in terms of, um, uh, so you mentioned something that in terms of housing and whether, um, uh, it can be used uh or whether that wealthy community housing should be or could be. Um, diverted into, um, into pension programs, uh, to be honest, I think, um, you know, this is a recent paper, and we, we, we find that result, but I'm not, um, you know, a, a real estate economist, and so I think, uh, I still need to learn a bit more about, um, what is out there in terms of why people, um, saving housing. I, I don't know if, if, uh, everything is out there, but, but certainly. Um, I, I, and perhaps we need to understand a little bit more why, uh, that's the case, why, you know, what, what, what roles, you know, housing is a very specific type of assets, and so what, what's its role, um, so I think in the literature, you know, that there's the consumption value of housing that is taken into account, but I don't know if the. Uh, maybe it can be used also to facilitate intragenerational transfers where you know, you, you have a big house and maybe you can reside with your children and in exchange for transfers. So all of that, um, I think needs a little bit more research to be understood to see whether in fact households are overinvesting in housing and because they, that's the only safe, uh, instrument that they have or, or because that's, uh, the, the optimal thing they can do. Um, Uh, and so I think, I mean you, you pointed out, uh, a lot of, um, of, of, um, of open questions on in terms of behavioral nudges and, and bundling and, and, uh, um, allowing for some short term access to these, these pension schemes and whether that would, would, uh, increase participation. I agree, so that's very, uh, very much, uh. Um, uh, the, the frontier and, and, and should be, should be researched. Uh, sometimes it's the difficulty with pensions is that it's hard to experiment, right, because once you start implementing a pension system in a certain way, you cannot really change it after, after two years easily. So that's, that's really, uh, the probably why, why we don't know more about these, these topics, but, um, that's why, you know, you can try to, to, to get at these questions uh in in direct ways as we've tried to do it in, in Pakistan. Um, and then, uh, moving on to Gustavo's remarks, um. On the, the disincentives, um, uh, about the minimum pension, I think in the talk, I talked, uh, you know, I, I talked mostly about the choice of that, that clawback rate, and what we find is that it doesn't, in terms of formality, it doesn't matter that much whether you claw back, um, uh, more aggressively or or less aggressively, uh, but that's not the whole story because, um, you know, the, I think one first order impact of minimum pension is what we call a wealth effect which is that once you start. Uh, transferring more money to people they're likely to retire earlier, um, and you know, to an extent that's kind of what we want, right? Like the minimum, the minimum pension is also there so that people don't have to work until they, they are completely, uh, um, disabled and, and unproductive, uh, but it's, it's to allow people to, to to retire at a decent age. So we also want to, to keep that in mind that, that, that's also what we're trying to achieve by having a minimum pension. Um, but yeah, so I, I guess my, the message of my research so far is that, you know, the, the, the, the, the, this implicit taxes consideration, they don't matter as much as you might think. Uh, really, what's going to matter is what level of, of consumption you want to guarantee, taking into account the fact that, yeah, that's going to allow people to retire earlier, uh, and, and, and maybe that's part of what we want to do. Um, vesting periods, uh, I haven't looked at that, uh, very much, even though it's, it's, um, uh, it, it, it's very interesting. I think Truman, uh, has, has worked on that, um, um, in my, my reading of the Chilean cases because in Chile they did have a vesting period for the minimum pension, but. Not that many people ended up right at that margin again, because uh you have a lot of, you have the people who work formally all the time and they will be way past the vesting period and then people who never work formally and they're, uh, they, they will never dream to, to reach that vest vesting period. So that reduces a little bit the margin that you're, that is at play when you're considering these disincentives. Um, um, and, um. And I think I forgot your last question, so uh maybe you can remind me, or, or otherwise, uh, yeah, I don't know if there's other questions. We have 4 questions in the room, so can I ask you to maybe just go quickly? Um, we have a little time, but, but, um, maybe start closest to me and then go around, so. Um, hi. Uh, my name is Priyanka. I'm with, uh, H&P. I actually moved from SP last year, and we're working, um, on social health insurance, um, and trying to understand, you know, some of these issues, um. With respect to willingness to pay, with respect to, you know, compliance, and I see a lot of similarities, um. In terms of, you know, pension pension contribution being You could use that as a proxy for compliance for estimating compliance, uh, minimum level of benefits, um, you know, the hyperbolic discounting that you speak about so one question is, um, whether you think we could use a similar framework for understanding, um. Some of these issues with respect to social health insurance and um I guess the other question is if if there is all this work happening on informality is is this also something that's being looked at and if so I'd love to know more about it thanks. Thanks so much and happy birthday Mihal. Um, my name is Katriel. Uh, maybe this is less important in light of the political economy considerations that you just highlighted, Clement, but, um. I'm curious how seriously. How confident we can be in the estimate of the effect of um contribution rates on informality. And I sort of see two subs questions that this is, I think, especially, I mean, it's especially interesting whether that translates into Um, affect the estimate, the true effect in, in poorer countries than Chile. Um, one sub question is like, what are the parameters that are really driving that? Is that the, is that the rate of time discount, sorry, time preference, um, or is it the other kinds of preferences that people have, um, and then. The The second question is just the robustness of that kind of that structural model. Do you have, uh, has it, does it make predictions, is it, has it made predictions that you've been able to validate on sort of out of sample data? Over. So my question follows uh a bit what Mihal um uh mentioned and I'm not sure to what extent this incorporated both in in research and in in the global practice and in in governments to what extent, uh, there is a uh it's when I mean you describe the issue of pension, but pension is usually bundled I mean it's associated with other benefits that workers might have when they, they join the workforce. Including maybe health insurance, uh, like, um, disability insurance, etc. and how much taxes they're, they're going to be paying and in and, and so I mean the decision to work on or not work, uh, in the formal sector is not just great to the pension. There is a lot of, um, things associated with that and also, uh, from the government side, uh, whether all those benefits are provided or not will affect the. Person's decision to work in the formal sector or not. So my question is to what extent both in the research and in the in in the practice governments have taken that into account how for example they are adjusting uh taxation, uh, income taxes or not income taxes, uh, that will affect the the decision of workers to both to work in the formal sector but also, uh, how much money they will have to pay for the pensions that they might be providing. Thanks, Miller. Sorry, I'm, I'm gonna call on Melis, and then, and then just after Melis I'm gonna call on Gonzalo who's got a hand raised, and then I'll turn it back to the presenters and then we'll conclude. So go ahead. Thank you so much, um, and thanks so much, uh, Clement, uh, for the presentation and as well as for the collaboration. I think, uh, with the work we've done together, uh, we managed to. Uh, operationalize, um, informal sector, um, savings, um, um, or, or pension scheme in a couple of countries, um, we are working on another one currently, so, so I think, uh, that's really important and to complement the work as, as Mihal mentioned, uh, we are carrying out some surveys to inform design. And I think the one in Pakistan um um maybe respond to some of the questions that um that uh are put on the table in terms of maybe why um you know informal sector might be investing in housing it's not a specific question as to housing, but you know there are questions related to trust in terms of where the informal sector um would be actually uh willing to and and be comfortable uh to to put their um money in. And I just also wanted to mention uh about the integrators right? uh it's real because um these integrators these um informal sector associations are being relied on significantly in Rwanda uh to expand coverage, uh, certainly in the MMA scheme in Kenya, uh, and, uh. There's another scheme just now which is called Habahaba in Kenya and they do also, um, rely on, uh, informal sector associations and and cooperatives as um as integrators. So, so they have been successful, uh, in those areas where I think, um, in a way the informal workers are are organized. Thank you. Thanks, Melis. Uh, Gonzalo, are you still there? Yes, thank you very much. Uh, thanks for the presentation. Uh, my question is about the, uh, well I have two comments about the, the, the knowledge gaps. Um, one is, uh, in terms of the the nature of the informal sector that I think is also changing and we have now, for example, the, the emergence of the gig economy. So how we can, um, also include that as a, as a feature in our, in our research, uh, and. If that open if that opens a new challenge or even gives us new opportunities to um uh to tap into to tap into the the possibilities of using the semi semi organized uh gig gig uh informal sector that is now integrated into the gig economy because they're organized through apps and this this type of work so that is a new, a new feature that gives a new opportunity. And the second is the um the 2nd, the 2nd is uh on the uh On the possibility of tapping into other sources of of of contributions from the informal sector, in particular what we have been thinking about is the informal sector or the informal income is not very visible, but all the. Informal workers, uh, they have to consume at some point and whether consumption will be more visible to tap into as a source of, of contributions. So that is another, uh, part of the research agenda that we can, we can expand. Thanks. Thanks, uh, thanks, Gonzalo. So I'm gonna let Clement maybe react and maybe conclude from your side, and then I'll close with, with Mikhael to close the session. Thanks. Um, thank you. Yes, um, all right, so Priyanka, yes, I, I agree, Priyanka, that, that a lot of the same questions, um. Or, um, that relates to participation in pen in pension schemes will be relevant for, for health insurance schemes even though, uh, often health insurance, you know, provides more immediate benefits so there's less of this question of present bias, etc. uh, but I think in both cases, as I said, and there's a lot of things we need to, to learn or document about the these non-poor informal sector that we're trying to incorporate into the schemes that will be useful for both, uh, uh, for both, um. Um, in terms of, um, of the, the effects that we find on the for the contribution rate and what, what's driving them, uh, I think yes, time preferences will be, will be, uh, will be a factor obviously, uh, because that's going to reduce the how much value those pension contributions, but also what I was trying to, to, to say which is that, that, um. You know, how far are people from the margin of indifference between formal and informal, so to the, you know, to what extent are, are people just um excluded from formal work because they don't have the right skills or because there's not enough jobs or, um, or what are the roles of these non-pecuniary preferences. So all of, all these are parameters in the model that, that, that play into the, the margin that we, we ended, end up measuring. Um, in, um, in terms of validating the, validating the, the results, so typically, so in those papers, um, we didn't incorporate, uh, exogenous variation that, that would, uh, would validate that, that, uh, that variation that those, uh, those numbers, but, uh, we, we drove, drew from the literature. So for example, in the, in the, the reform, the 2008 reform, uh, was, um, analyzed also with reduced form methodologies that. That exploit how different birth cohorts are are exposed differently to, to the change in rules and, and the magnitudes uh that we find are, are similar um but of course what we can do then with the structural model is is simulate designs that did not, um, that did not, uh, uh, that were not implemented. Um, And Sergio, uh, I'm sorry, I'm, I'm not sure I understood, uh, your question very well. So how much are government taking into account this, the evasion in terms of setting their, their, to what extent the decisions of pension are related to the decisions on, on taxes or provision of, of health insurance, disability insurance, etc. To what extent this is isolated or is viewed as a package or because I mean they are all related. Yeah, I mean, theoretically, as, as I think they're, they're, they're, they're, they're related even though how much value your contribution to a pension versus your contribution to uh health insurance is going to be different. For example, if you're very present biased, you might not value pension contributions, but you might value the. Uh, contribution to health insurance, um, and, and, and again taxes is the same, right? Like there's, uh, or you know if you're completely present biased and your pension contribution you're gonna value it, uh, as a tax, exactly as a tax, right? So, uh, from the government side whether they take the, the government takes, uh, the decision to pension in an isolated way relative to other, other provisions for social safety net or, or taxes. Yeah, yeah, I'm not sure. I, I, yeah, we can, we can talk we can talk, uh, one on one, and, um, Meliss, thanks for, for, uh, I mean it's always great to talk to you because you have all these, uh, this knowledge of, of what's going on everywhere, um, and, and, um, and yes, um, uh, in terms of informal sector, uh, professional organizations, um, what I was, I agree there's a lot of anecdotal evidence. That governments are are interacting with these these professional organizations in many ways and practitioners know about it but what we don't have is like some systematic uh way of documenting what how many let just if if only what fraction of the informal sector is affiliated to an organization so that's where I think research needs to happen to to to make it more systematic, um. And lastly, um, I'm not sure I have much to say about like the, the gig economy, so I agree that, that, that it, it poses a lot of similar questions that, um, um, to, to, to, to what we've been discussing about, uh, informality, um. In, in terms of, uh, if financing pensions with, with uh let's say tax on, on consumption, I think, uh, Pierre, uh, Pierre Bachas in the research group has research that looks at what fraction of consumption is informal or is, is, is done, uh, in informal means and therefore would escape taxation. And so I think that's a very important parameter to take into account in order to understand who would be, uh, impacted by, by, uh, taxes on consumption that would try to finance pensions, um, so yeah, I think it's a very interesting, uh, dimension to look at. Great. So Miguel, you, you, you have the floor for to close. Thank you. I have 4 points to make before, before I turn back to you to close. But, uh, the first is on Prianka, what Prianka was saying, which I, I think it is extremely important. Uh, so yes, there is a lot of, there are a lot of analogies between social health insurance and pension insurance, except that I would still insist on a distinction between. Retirement pensions and all other kinds of pensions like disability survivorship. What is the distinction? The distinction is that in the case of survivorship or disability, there is a huge analogy with health insurance. This is really a part of social insurance. Um, and, and we can talk about similarities. Obviously maybe myopia is smaller in health, but, but then there is this big issue of creating the health package. I see retirement pensions as they look like in the world slightly differently, primarily because it stopped being an insurance. You can think reasonably about insurance if the probability of something occurring to you is, I don't know, it was like 100 years ago receiving retirement pension. The age of eligibility for retirement, 1 in 6 workers was getting it, 1 in 8. Now we have a situation where 9.5 in 10 get it, so it doesn't lend itself to insurance type thinking. It lends itself to savings type thinking. This allocation of consumption over the lifetime is fairly predictable. That's why I am a big fan of defined contribution pension schemes, but I'm talking retirement pensions, totally different ball game with other sorts of pensions. So I want to make this distinction which I think makes it different than health. In terms of retirement pension, which of course we need to look at what income level, uh, be that, that, that becomes blurred when we talk about formalizing the informal sector, but certainly we need to do work together and I always wanted us in the bank to do more work together in the regions and globally, and it hasn't happened like, you know, uh, from our GP folks from HNP from DC sitting together and doing things that would be, we, I don't think we've done enough, enough of that. Second point, um. Uh, second point, uh, on, on, uh, taxation and social insurance again, the situation that we haven't done enough and that's why it is so exciting what is happening in luck now because this is essentially looking at tax and benefit reforms so that to improve redistribution and risk sharing in social protection schemes. So in this sense this, this, this is precisely when this is going to be brought together as we will be discussing this next week. So I'm very excited. About it, two last points. One is Gustavo talked about vesting period that may be a problem, but Clement, we discussed a few days that the lack of vesting period may be a problem because I didn't know that in Chile, even if you have a very small savings, it's still being converted into some kind of annuity which makes it a mockery because it's so low. So it would be better if this possibility didn't exist, meaning unless you accumulate something which is in part of the. Resting period, right, unless you accumulate through that, then you cannot get a pension. The moment you anyway it's for separate discussion what's what's better, what's the right vesting period. But I understand too long of a vesting period is a problem, but the lack of a vesting period is another problem with different consequences. Finally, my final point is that the way I got Gustavo's comment was, was maybe I am now putting my words in his mouth. It's about observability. Of transactions, given that so many transactions are observable transactions in terms of consumption, because mobile money is used, could we use it as a tax base for social protection purposes by automatic default that $1 goes somewhere? If you do a transaction like I heard, I haven't seen it, that in Ethiopia, not Ethiopia, in Estonia, in cafeterias in Estonia, you have schemes when you save $1 automatically as a default if you buy coffee. Because it's observable and you use your, your card, of course you can opt out, but the point is the default is there. So and you can choose whether you want to be in this. I don't know the details, but I think that line of thinking, uh, in addition to thinking about what's the right subsidy rate from general revenue to contributions, that's one thing, but the other thing it could we automatically be taxing transactions. And then earmark or not to earmark is yet separate discussion, but I think we see it often in our discussions as a discussion that has big potential for coverage of the informal sector. And thank you very much again for having me here. Really great birthday gift. So much appreciated. Thanks, Michel. Thanks, Clema. Thanks to all of you for great questions and, and, uh, we'll see you next month for our next PRT. Thanks a lot.

showAllTranscripts

no

duration

PT1H28M45S

scene7File

worldbank/PensionsAndInformality

scene7Domain

scene7FileAvs

worldbank/PensionsAndInformality-AVS

title

PensionsAndInformality

description

PensionsAndInformality

showTimestampAndTranscript

yes