00:03 For coming for the session,
00:04 uh,
00:05 so I,
00:05 I must thank Farhahlpi for asking me to moderate the session.
00:09 Uh,
00:10 she asked me because I think she wants me to learn
00:13 about,
00:14 about climate resilience.
00:16 Um,
00:17 so,
00:18 uh,
00:18 the,
00:18 the,
00:18 the,
00:19 the,
00:19 the way that I would introduce the session is that natural
00:21 disasters that are due to climate change and other things,
00:25 they cost the world about a quarter of a trillion dollars each year,
00:29 OK.
00:30 And they also cause a lot of debts,
00:32 uh,
00:33 but most of those debts,
00:34 about 80% of them
00:36 are in low and lower middle income economies,
00:39 uh,
00:40 so we think that the severity and the frequency of
00:43 these shocks has been increasing,
00:46 and
00:46 the main question,
00:47 and that's a question for the session today,
00:49 is
00:50 how can governments
00:52 and markets make
00:54 people and enterprises,
00:56 uh,
00:57 more,
00:58 uh.
00:59 Uh,
01:01 much better at handling these shocks in the sense of actually
01:05 both not being affected them nearly as much and to
01:08 the extent that they are affected to actually rebound quicker,
01:11 right?
01:12 So
01:13 to discuss this question we have uh 3 experts and I'm going to introduce them very,
01:17 very quickly and I introduce them all at once.
01:20 Uh,
01:21 so the first one is Michael Carter.
01:22 He'll go first
01:24 and he will discuss
01:25 whether insurance markets
01:27 can build resilience of households,
01:29 OK.
01:31 So I should say a few words about Michael.
01:33 Uh,
01:33 he is distinguished professor for agricultural and Resource
01:36 Economics at the University of California Davis.
01:39 He's done a lot of work in this area
01:41 and he's going to,
01:42 uh,
01:43 uh,
01:44 he's going to share some of that with us.
01:46 The second speaker is Matt Kahn,
01:49 and he'll talk about the interactions between private investment
01:53 and government protection against climate risk.
01:56 Matt is the Provost professor of economics
01:59 at the University of Southern California.
02:01 And the third person is Eko Swehi.
02:04 Uh,
02:05 uh,
02:07 she will go 3rd,
02:08 and she will discuss how the private sector,
02:10 uh,
02:11 can increase the financial protection
02:15 for people in vulnerable countries.
02:17 Uh,
02:18 so Ekoswehi is actually,
02:19 uh,
02:20 very well qualified to do this.
02:22 Uh,
02:22 she's been involved in two innovative,
02:25 uh,
02:25 2 innovative institutions and initiatives.
02:28 The first one is the African Risk Capacity Agency.
02:31 And the other one is the Caribbean Catastrophe
02:34 Risk Insurance Facility.
02:36 She's the general secretary of the Insurance Development Forum,
02:40 which is supported by the UN and the World Bank.
02:43 So
02:44 with those introductions,
02:45 I'm going to ask Michael to start us off.
02:47 Uh,
02:47 he'll speak for between 15 and 20 minutes.
02:52 Sorry.
02:55 There you go,
02:55 Mickey.
02:58 Thank you very much.
03:01 So,
03:01 good afternoon.
03:02 It's a great honor to be here.
03:03 I'm glad to be,
03:05 uh,
03:05 speaking,
03:06 uh,
03:06 with all of you and,
03:08 uh,
03:09 and,
03:09 and on this very important topic.
03:10 So.
03:11 Uh,
03:12 the slide you see in front of you,
03:13 those are women in,
03:15 uh,
03:15 northern Kenya,
03:16 specifically Samburu County,
03:18 who
03:19 have been involved in a graduation program study we did that also combined
03:23 asset building with asset protection.
03:26 Um,
03:26 so we'll talk uh perhaps a little bit about that.
03:30 I'm really gonna,
03:30 I'm doing something slightly unwise.
03:33 Uh,
03:33 I was told by a Jesuit priest many years ago,
03:36 never try to give more than 3 messages in a single talk.
03:39 I'm actually gonna try to mash together 2 papers,
03:42 each of which has about 12 messages.
03:44 So I'm,
03:45 I'm going against,
03:46 uh,
03:46 use,
03:46 I'm going against my training,
03:48 uh,
03:48 by trying to do this,
03:49 but
03:49 hopefully you'll see why I'm,
03:51 uh,
03:51 mashing together,
03:52 uh,
03:52 these things.
03:53 Let me give a small,
03:54 uh,
03:54 amount of credit.
03:55 One paper,
03:56 uh,
03:57 is with Sarah Jansen and Nobu Ik Kagami,
03:59 and it,
03:59 it asked the question,
04:00 if insurance can change,
04:02 uh,
04:02 poverty dynamics and reduce the cost of social protection.
04:05 I'm gonna draw on just a few ideas on that.
04:08 And then the second piece of work is with,
04:10 uh,
04:10 Moa Lus and it's on,
04:12 on measuring resilience with well-defined counterfactuals.
04:15 And so I'm gonna try to make a,
04:17 a transition between,
04:18 uh,
04:18 these two things in a way that,
04:20 uh,
04:20 hopefully will prove to be,
04:22 uh,
04:22 sensible.
04:23 Uh,
04:23 so real quickly,
04:24 let me just
04:26 summarize a little bit of theory of insurance markets and the cost of,
04:29 of social protection.
04:31 So in this paper I mentioned with.
04:32 Uh,
04:33 with Sarah and,
04:33 and Nobu,
04:34 it's a,
04:34 it's a calibrated,
04:36 fairly complicated dynamic stochastic programming model.
04:39 Model assumes households are very smart,
04:41 very forward looking,
04:43 understand probability distributions,
04:45 understand how insurance works,
04:47 um,
04:48 and they,
04:48 they do the best they can in a,
04:50 in a rather difficult environment.
04:52 That model is actually calibrated on
04:54 the area of northern Kenya where we just saw those women,
04:58 uh,
04:58 a moment ago.
04:59 Um,
05:00 so
05:00 what happens in,
05:01 in this model is,
05:03 as,
05:03 as we'll see in a moment,
05:05 but
05:06 the provisioning of insurance or inducing an insurance
05:08 market and specifically an index insurance market,
05:12 uh,
05:12 for people like those women
05:13 actually has substantial impacts
05:15 on,
05:16 on poverty dynamics,
05:17 and it,
05:17 the,
05:17 uh,
05:18 the,
05:18 it has effects through three mechanisms.
05:21 The first is what I call a vulnerability reduction or resilience effect,
05:25 so it,
05:26 it keeps people from falling as far.
05:28 And so if you don't follow as far,
05:30 it doesn't take you as long obviously to get back.
05:32 Uh,
05:33 the second,
05:33 it changes investment incentives for people,
05:36 so it actually has what I call a,
05:38 uh,
05:38 a resilience dividend effect.
05:40 So we actually
05:41 alter the long term
05:43 level of poverty because,
05:45 uh,
05:45 providing insurance for people's assets actually
05:48 enhances their incentives to accumulate and wisely invest.
05:52 And the third,
05:52 which is sort of a strange one,
05:54 and when I first introduced this a few years back to Alain de Genvri at Berkeley,
05:58 he said.
05:58 You're making people pay for their own social protection,
06:01 which is true,
06:02 uh,
06:02 so in this model we're gonna look at the wisdom
06:05 of a partial government subsidy program for insurance,
06:09 um,
06:09 and that actually allows the existing government budget,
06:13 uh,
06:13 to stretch further.
06:14 So we'll get,
06:14 I'll,
06:14 I'll just summarize those results,
06:16 uh,
06:17 really,
06:17 uh,
06:18 really quickly,
06:19 and they can be summarized.
06:20 I think I've been seeing everyone say that this pointer doesn't work.
06:23 So
06:23 what we can see in these two diagrams,
06:25 I don't need to go them in,
06:26 in great detail.
06:27 But the diagram on the left is a poverty headcount that comes out of this,
06:31 uh,
06:32 this,
06:32 this long-term model over a period of 50 years.
06:35 The one on the right is the poverty gap.
06:38 The solid,
06:39 I guess it looks black line,
06:40 the uppermost line in the latter part of the period.
06:43 That's if the government,
06:44 uh,
06:45 if there's no insurance market active.
06:47 The sort of bluish
06:49 line,
06:49 which is in the,
06:50 in the middle in most of them,
06:52 is when there's an insurance market,
06:54 but the government leaves it up to people's own
06:57 devices,
06:57 whether or not they wish to purchase insurance,
07:00 and the red line is sort of a sloppily targeted.
07:02 Subsidy program which the government offers to to bring subsidies
07:06 and so what you can see from this
07:08 is it,
07:09 it really does in the in the context of this model,
07:11 it really does alter
07:13 long term poverty dynamics so
07:15 you can see the headcount measures of poverty
07:17 end up being much lower
07:19 uh when you have an insurance market presence
07:22 uh and also the poverty gap eventually
07:24 uh becomes smaller.
07:25 There's some interesting intertemporal trade-offs there which I,
07:28 we can talk about if,
07:29 if there's time.
07:30 Uh,
07:31 or interest,
07:32 the,
07:33 the main impacts,
07:34 the largest growth there are is actually coming
07:36 from what I call the vulnerability reduction effect,
07:39 OK,
07:40 uh,
07:40 and,
07:40 and then the red line,
07:41 which is when the subsidy comes in,
07:43 uh,
07:44 is the additional investment incentive effect.
07:46 So one of the things that comes out of this work
07:49 is if you are in a dynamic system
07:52 that has a tipping point or a poverty trap,
07:54 if you are vulnerable in the sense that you're close to that dynamic tipping point,
07:58 essentially the shaft.
07:59 The price of money
08:00 is extremely high for you,
08:02 and it turns out it's not optimal for people
08:04 right at the tipping point to buy any insurance.
08:07 It's not that it wouldn't be valuable,
08:09 but the money is,
08:09 it's actually more valuable for them to get as far away
08:12 from the tipping point as quickly as they can.
08:15 When you've subsidize insurance,
08:17 it takes those vulnerable,
08:18 those most vulnerable people,
08:20 and then they,
08:20 then they're,
08:21 they're willing to pay sort of half price on it.
08:23 And so that's where you get some additional poverty reduction effects.
08:27 Uh,
08:27 the,
08:27 the,
08:28 the final kind of exercise we do in the paper is we,
08:31 we provide the following thought experiment.
08:33 Suppose the government comes along
08:35 with a social compact and says,
08:37 let's guarantee
08:38 everyone in this economy
08:40 that we're gonna raise them up to the poverty line,
08:42 OK?
08:43 And the,
08:43 the black solid line now is the budgetary implication
08:47 of the government
08:48 holding on to that social compact
08:50 when there is no insurance market.
08:53 You can see over time the black line
08:55 uh sort of creeps up as government expenditures begin to increase.
08:59 And then we ask ourselves,
09:00 is there really a public finance case
09:03 for
09:03 inducing insurance markets,
09:05 and secondly for subsidizing those insurance markets,
09:08 given the particular behavior
09:10 of the the particular
09:12 behavior of people who are
09:14 in this system
09:15 with a dynamic tipping point.
09:17 Uh,
09:18 when we do the comparisons here,
09:19 let me just focus on the subsidized case,
09:21 that's the red line again,
09:23 um.
09:24 When we do that,
09:25 we,
09:25 we,
09:26 we,
09:26 we measure total government expenditures as closing the poverty gap for everybody
09:30 and also paying for the insurance subsidies.
09:33 So you can see initially,
09:34 the cost
09:35 of that program is rather high,
09:38 but by the time you get to the end of the fifty-year.
09:40 Uh,
09:41 simulation because you've had these investment incentive effects,
09:44 you've reduced vulnerability for everybody.
09:46 Government expenditures are actually less,
09:49 and the paper goes into much greater detail,
09:51 but
09:51 there on the slide you can see that if we take the discounted present value of the.
09:56 Different kind of public expenditure streams,
09:59 the government actually saves money
10:01 even by subsidizing insurance
10:03 because you're basically,
10:05 you're keeping more of the population
10:07 in a,
10:07 in a circumstance where they can
10:10 maintain their economic viability.
10:12 So as I often,
10:13 I think as as the paper actually puts it,
10:16 um,
10:16 we can,
10:17 we can,
10:17 uh,
10:18 we,
10:19 we can imagine a world
10:21 in which.
10:22 Uh,
10:23 you know,
10:23 we might say insurance subsidies are expensive,
10:25 but this is also saying
10:26 not subsidizing insurance is also expensive in,
10:29 in terms of the poverty that,
10:31 that emerges without
10:33 sort of good social protection mechanisms,
10:35 um.
10:37 Something I'll just very briefly flag here,
10:40 we're looking at insurance mechanisms here to achieve this goal.
10:43 It works because people understand and trust the insurance.
10:46 It's reliable.
10:47 It's delivered quickly.
10:48 In principle,
10:49 you can have governments do shock responsive
10:51 social protection programs.
10:53 Uh,
10:53 government of Kenya is experimenting with it
10:55 with something called HSMP 2,
10:57 which is a scalable
10:59 social protection program.
11:01 What I've heard about how that works,
11:02 doing a lot of field work in the area,
11:04 we get a lot of,
11:04 uh,
11:05 rural myths about how that works.
11:07 So I,
11:07 I would say there's not a lot yet that tells us that that works,
11:10 uh,
11:11 necessarily very well.
11:12 Um,
11:13 there's another side thing since we have someone who worked for ARC before,
11:17 um.
11:17 You know,
11:18 there's a side issue if the government is in fact gonna carry that risk
11:21 of stochastic social protection obligation,
11:24 should they in fact
11:25 engage with the Africa risk capacity,
11:26 for example.
11:27 Uh,
11:28 we've been doing some work on when that is good public finance,
11:31 uh,
11:31 or not,
11:32 and that's a,
11:32 that's again a side issue,
11:33 but I think it's actually a really,
11:35 uh,
11:35 important question to consider.
11:37 OK,
11:38 so
11:38 theory is very suggestive,
11:40 right?
11:41 We,
11:41 we've just seen a couple of ways
11:43 through,
11:44 through which,
11:45 uh,
11:45 insurance could actually or insurance markets,
11:48 perhaps subsidized insurance markets could
11:50 increase resilience,
11:51 but if I stop to think about it,
11:53 I realize
11:54 I don't actually know what resilience is.
11:56 I don't think there's any good measures.
11:58 Uh,
11:58 in fact,
11:59 uh,
12:00 excuse me,
12:01 yesterday,
12:01 in Dermit,
12:02 in your introductory remarks,
12:03 you gave,
12:04 you gave a definition of resilience,
12:06 which is pretty close
12:08 to that first bullet point on there,
12:10 right?
12:10 I think
12:11 you said,
12:12 well,
12:12 resilience is kind of when you get back to where you were pretty quickly,
12:15 OK.
12:16 Um,
12:16 so I wanna suggest that we take that definition of resilience seriously
12:20 and figure out how to do it
12:22 because,
12:23 uh,
12:23 I was sort of doing a resilience like analysis in that prior paper,
12:27 but I wanna be a little more formal
12:28 and I wanna I wanna sort of think about it.
12:30 So again.
12:31 If,
12:32 if we really wanna measure
12:34 resilience as,
12:35 as,
12:35 as I think we all really mean it when we talk about economic resilience,
12:39 then obviously we're gonna need to observe
12:42 whatever unit of analysis like households that we're interested in,
12:45 we're gonna need to observe them.
12:47 After a shock,
12:48 right?
12:48 And more importantly,
12:49 we're gonna need a counterfactual estimate of where
12:52 they would have been without the shock,
12:54 right?
12:55 Because if we wanna measure
12:56 the cost of the shock,
12:57 if we say being resilient means we have
13:00 very small loss of current and future well-being,
13:02 then we need a counterfactual if we wanna measure what that would have been.
13:07 So what I'm gonna share with you quickly then
13:09 is,
13:09 is,
13:10 is taking a look.
13:12 At,
13:12 at
13:13 first,
13:13 how we can define a measure,
13:15 I'm gonna do it with fake data because it's always,
13:17 it's a lot easier to shock people inside your computer CPU
13:20 than it is to do it in the real world,
13:22 thank God,
13:23 uh,
13:24 and,
13:24 and then I'm actually gonna use some real data and show you how you can use it
13:28 and,
13:28 and bring it around,
13:29 uh,
13:29 bring both examples if time permits
13:32 to,
13:32 uh,
13:33 uh,
13:33 to this,
13:34 uh,
13:35 to,
13:35 uh,
13:35 to,
13:35 to insurance and,
13:36 and its impacts on resilience more formally,
13:39 uh,
13:40 defined.
13:40 So I'm just gonna jump ahead here a little bit.
13:43 Um.
13:45 So the upper,
13:45 the upper curve there
13:47 is the expected value of income
13:49 for a
13:51 control group that does not receive a shock.
13:53 The lower line,
13:54 so these are people who are behaving dynamically optimally.
13:57 So you can see the control group is,
13:59 is on an upward trajectory converging
14:01 to sort of their solo steady state if you want to think about it,
14:04 uh,
14:04 that way,
14:05 and that's more or less the right way to think about it.
14:08 We take a random sample of the people in our study,
14:11 and we give them a horrific shock,
14:13 OK?
14:14 So they immediately have a very large collapse.
14:16 And then the red line,
14:17 as it slowly recovers,
14:19 right,
14:19 that's actual optimal rate of recovery when people
14:22 are considering in the absence of ability to borrow,
14:25 that's how fast they will
14:27 try to return to their counterfactual state.
14:30 OK?
14:31 So this is made up with fake data.
14:32 I skipped over the slides that say how we estimate it,
14:34 but it's,
14:35 it's pretty obvious.
14:36 Uh,
14:37 and there's a very good literature,
14:38 as I think we all know in this,
14:39 uh,
14:40 in this room that looks at the,
14:41 the impact of shocks on,
14:43 on coping,
14:44 on investment,
14:44 and all,
14:45 all sorts of things.
14:45 So this work really
14:47 draws on that,
14:48 on that tradition.
14:49 Moe and I in this paper suggest we define resilience this way just graphically.
14:54 Uh,
14:54 resilience is defined there in the middle.
14:57 L is the gap
14:58 between,
14:58 it's the loss gap between
15:01 the actual income level of people who were shocked
15:03 and their counterfactual measure of where they would have been.
15:07 OK,
15:07 so that's the core part of the measure.
15:10 And then just to normalize it.
15:12 Then we basically normalize that loss gap
15:16 by
15:16 an imaginary standard if they didn't recover at all.
15:19 So if they didn't recover at all,
15:21 they would,
15:22 they would go along the bottom of that gray
15:23 box at what we call the zero resilience line.
15:26 So defining it with this normalization is,
15:28 is kind of nice,
15:30 um,
15:31 because it makes the resilience measure go between 0 and 1.
15:34 So if you don't recover at all,
15:35 you have zero resilience,
15:37 and if you were to recover instantaneously,
15:40 you would have a resilience
15:41 of 1.
15:42 OK?
15:43 In this particular instance that comes from our dynamic programming model,
15:47 uh,
15:47 the resilience measure is actually 43%,
15:50 and you will note that's cutting off the measure,
15:52 uh,
15:53 uh,
15:53 six years after the shock occurred,
15:56 OK?
15:56 Uh,
15:57 so that's just an example
15:58 that you can take.
15:59 Imagine having panel data like this.
16:01 You could actually measure
16:03 a shock,
16:03 you know,
16:04 subject to some identification
16:06 concerns.
16:07 I'll just mention quickly,
16:08 you can actually also measure it at the individual level
16:12 and so I won't explain the,
16:13 the mathematical expression there,
16:15 but that's just a matching estimator for individual resilience.
16:19 It's the same concept
16:20 of comparing
16:21 how.
16:21 Household eye with their counterfactual estimate of where they were
16:24 and we can,
16:25 we can add up their individual resilience.
16:28 I think that's really important if we measure
16:30 resilience
16:31 because then we can start to look at the determinants of resilience,
16:34 which I think is something that,
16:36 that,
16:36 that we need to,
16:37 uh,
16:37 kind of get to.
16:39 Another reason for doing this,
16:40 it,
16:41 it,
16:41 it allows us to think about
16:43 the,
16:44 the cost-benefit analysis or more generally the economic value
16:47 of resilience promoting interventions.
16:50 So first,
16:50 let me real quickly
16:51 take an example from the computer.
16:53 So,
16:54 uh,
16:54 we imagine,
16:55 uh,
16:56 uh,
16:56 uh,
16:56 a catastrophic insurance policy
16:59 is introduced.
16:59 It only pays off if the shock is really bad,
17:02 OK?
17:03 And we and and and it's calibrated to cover
17:06 1 in 10 year events.
17:08 So to make the analysis kind of balanced,
17:10 we assume that someone's buying this insurance for 10 years
17:14 before the shock occurs.
17:16 And then when the shock after the shock occurs,
17:18 the green line there
17:20 shows the much more rapid recovery
17:22 of households that had their assets
17:24 kind of rebuilt or partially rebuilt subject to a deductible,
17:27 OK,
17:28 and.
17:28 And once you do that,
17:30 the,
17:30 the resilience measure goes up quite substantially.
17:33 Um.
17:35 In the paper we,
17:36 we look at two cases.
17:37 I'm actually sharing with you the non-poverty trap,
17:39 uh,
17:40 paper.
17:41 Uh,
17:41 we can see when we,
17:42 when we provide this catastrophic insurance,
17:45 resilience actually rises to about 70%,
17:48 so this resilience measure goes up substantially.
17:50 If we do a benefit cost analysis and we basically
17:54 take the present value of that area in the middle that's marked Delta.
17:58 OK,
17:58 and we compare that to the the present value of that to the
18:01 present value of,
18:02 of having made insurance payments
18:04 for 10 years,
18:05 you can see the benefit cost ratio is about 1,
18:08 just under 2,
18:09 OK.
18:10 This kind of approach actually accommodates poverty traps.
18:13 I was talking to
18:14 Hebram today wherever he,
18:15 I saw him a moment ago,
18:16 um,
18:17 who's thinking about vulnerability,
18:19 which
18:20 in a sense
18:21 these upper measures are actually perfect measures of vulnerability as,
18:24 as well.
18:25 Um,
18:26 OK,
18:26 let's go real quickly to the,
18:28 to the real world to sort of
18:29 wrap up.
18:30 Um,
18:31 There's a,
18:32 there's,
18:33 if you look at the slides,
18:34 there's a,
18:35 it tells you where to get the different papers here.
18:36 Um,
18:37 and if,
18:38 if God smiles on us,
18:39 this boucher at all will be published before my next birthday.
18:42 Now,
18:43 probably not.
18:43 My birthday is in 2 weeks,
18:44 so
18:45 there's no,
18:46 there's no hope because of referee 4
18:48 is doing it to us again.
18:50 Uh,
18:50 this was a 4 year
18:52 RCT
18:53 in Tanzania,
18:54 Mozambique.
18:55 Uh,
18:55 we were looking at bundling stress tolerance,
18:58 specifically drought tolerant seeds,
19:00 with an insurance contract
19:02 which was intended to cover the risks
19:05 that
19:06 drought tolerant seeds don't.
19:07 I was talking to Chave,
19:08 uh,
19:09 earlier today
19:10 because he said,
19:10 why do you need insurance if you have drought tolerant seeds
19:13 and the reality is drought tolerant seeds
19:15 or even flood tolerant rice only cover a subset of the events that happened.
19:19 So we came up with an index insurance contract
19:22 which I could talk about for hours,
19:23 but I will not.
19:24 Um,
19:25 and,
19:25 and then we end up having a treatment.
19:27 We have a control group.
19:28 This was a spatially diversified RCT,
19:31 so we had two countries,
19:32 Mozambique,
19:33 Tanzania.
19:34 We spent a lot of time looking at weather maps,
19:36 so we tried to maximize the probability
19:39 that somebody somewhere in our study would get,
19:41 would get zapped,
19:43 uh,
19:43 which is an unfortunate way to design a study,
19:45 but that's what we did.
19:46 Uh,
19:47 nature,
19:48 fortunately for us,
19:49 unfortunately for the people in the study,
19:51 the nature was pretty generous in terms of providing
19:54 different kinds of shocks,
19:56 both the kind of shock that is managed,
19:59 um,
19:59 that is managed by the drought tolerant seeds,
20:01 as well as the kind of shock that's covered
20:03 by the insurance.
20:05 Uh,
20:05 I can skip,
20:06 this is junkie,
20:07 we'll get over it.
20:07 I mean,
20:07 that's just saying we regress.
20:09 We're just regressing,
20:11 uh,
20:11 various kinds of outcome measures.
20:13 I'm gonna look at May's income here,
20:14 regressing it on the shocks and the interaction between the shocks
20:18 in the different,
20:18 in the different treatments and controlling.
20:20 For the usual stuff
20:22 you would expect to see controlled for in an RCT.
20:24 OK,
20:25 so now this is real data,
20:26 OK?
20:27 So now the counterfactual line,
20:28 the upper blue line that we're looking at,
20:30 this is what happens
20:32 to our control group,
20:34 who does not receive a shock.
20:36 OK?
20:37 And you can see the importance of the counterfactual here.
20:39 So,
20:39 uh,
20:40 what's labeled there is year 3 was actually not a great year.
20:43 So even the people that weren't shocked were sort of
20:45 not doing great in year 3.
20:47 And then the red line is actually the group of people
20:50 who were hit by severe shock in year 2.
20:54 OK?
20:54 And what we see is
20:56 there's extremely low resilience,
20:58 OK.
20:58 By the following year,
21:00 these families have barely recovered.
21:02 If
21:03 the paper breaks it down
21:04 into drops in investment at the intensive and extensive margins,
21:08 kids don't eat,
21:09 all sorts of terrible stuff is going on there.
21:11 OK,
21:11 but it's sort of summarized here
21:13 as a lack of resilience.
21:14 And then we,
21:15 we sort of ran out of money,
21:16 so the study came to an end,
21:18 but we sort of imagined they would slowly
21:20 start to recover.
21:21 Uh,
21:21 but the way I've written it here,
21:23 the measured resilience is only about 18%.
21:26 OK,
21:27 so there's a lot that could happen.
21:29 Now,
21:29 what happens to our
21:31 experimental treatment group?
21:32 OK,
21:32 so again,
21:33 we have an experimental randomly selected treatment group
21:36 that receives a package of risk mitigation
21:38 technologies,
21:40 OK?
21:40 And so a couple of things are going on here.
21:42 So that's the green line.
21:44 OK,
21:44 and again this is all based on the econometric,
21:46 uh,
21:46 estimates.
21:47 And
21:48 at year two,
21:49 we give the identical shock.
21:52 To the people in the treatment group,
21:53 OK.
21:54 The first thing you notice is they don't fall nearly as far.
21:57 That's because it turns out
21:59 that drought tolerant seeds actually work.
22:01 I actually did not believe they would.
22:03 I thought they only worked on fancy farms with,
22:05 with,
22:05 uh,
22:05 fancy agronomy.
22:07 Turns out they work for like farmers that have pathetic soils,
22:10 use no inputs other than seeds.
22:12 So that's a really interesting finding.
22:14 So first of all,
22:15 they don't fall very far,
22:16 and
22:17 more importantly,
22:18 they're not,
22:18 they're not capitalized for the next year.
22:21 OK,
22:22 so
22:22 for the
22:23 control group farmers that were shocked,
22:26 their input intensity drops tremendously,
22:29 and they actually reduced the area planted
22:31 by about 25% the year following a severe shock.
22:34 So the,
22:34 the household is really hunkered down
22:36 as best they can,
22:37 and they're not at all resilient.
22:39 Um,
22:40 so if we ignore the part for a moment that's labeled
22:43 resilience dividend above the blue line,
22:45 what we find is
22:47 this
22:48 this treatment group has almost perfect resilience,
22:51 and it's a combination
22:52 of the action of,
22:53 of a genetic technology as well as
22:56 index insurance,
22:57 uh,
22:57 kind of financial,
22:59 uh,
22:59 technology.
23:00 The final thing,
23:01 and,
23:02 and I'll kind of end here since the,
23:03 the,
23:04 the buzzer just is going off.
23:06 The final thing,
23:07 are we OK?
23:09 Absolutely fine.
23:09 It's an excellent.
23:10 Oh,
23:10 OK.
23:11 Well,
23:11 thank you.
23:13 Um,
23:15 So just,
23:16 just real quickly,
23:17 the really interesting thing is the year after the shock,
23:21 we see the so-called resilience dividend.
23:23 So those of you who have looked at the literature on insurance,
23:26 at least when it works,
23:28 right,
23:28 there's this ex ante or resilience dividend effect.
23:31 So when people become convinced
23:33 that they have a financial or in this case
23:35 a financial slash agronomic technology
23:38 that will
23:39 help them,
23:40 they actually intensify.
23:41 Investment.
23:42 So what's going on there with that big post-shock jump is
23:46 not only did these mechanisms keep them from falling very far,
23:49 not only did it help the insurance help them preserve their capital,
23:52 but it seemed to have given them confidence,
23:55 and they actually expanded
23:56 the year after a shock
23:58 when everybody else is shrinking
24:00 their production.
24:01 They actually expanded their production from
24:04 2 to like 2.6 hectares.
24:07 So it's a huge jump,
24:08 and they also intensify the amount of inputs that they use.
24:11 OK.
24:12 Now the flip side of this in the paper
24:14 is being tortured by referee 4 who wants us to prove for sure
24:18 that this was a learning effect,
24:20 and we think we've satisfied
24:22 my favorite referee 4.
24:25 but what we think,
24:25 you know,
24:26 if you look at the treatment group
24:28 people
24:29 who did not receive shocks.
24:33 In this year 3,
24:34 they actually run away from the technology,
24:37 and their behavior
24:38 and productivity
24:39 becomes indistinguishable
24:41 from the control group.
24:43 OK.
24:43 So to me this is an example of farmers,
24:46 if a bunch of people come up in,
24:47 in,
24:48 in trucks from cement or wherever our trucks came from,
24:51 you know,
24:51 and they say,
24:52 oh try these magic seeds,
24:53 and they think
24:54 Jack and the Beanstalk,
24:55 my mother warned me about this,
24:56 right?
24:57 But
24:57 most people seem pretty accommodating.
24:59 And they,
25:00 they sort of dabbled in it,
25:01 right?
25:02 And after the first year,
25:03 people that didn't see the benefits,
25:06 uh,
25:07 uh,
25:07 you know,
25:07 walked away from it.
25:09 People that did see the benefits says,
25:10 wow,
25:11 this is amazing,
25:12 and they intensify.
25:13 I think there are huge policy lessons there.
25:16 I've written a little bit actually with Alanda Jadri and Jing Kai,
25:19 who's at Maryland on how do you subsidize
25:22 an input
25:23 which
25:24 delivers its benefits only occasionally.
25:27 Right,
25:27 because people,
25:28 you know,
25:28 the expression I learned as a kid,
25:31 uh,
25:31 was,
25:31 you know,
25:32 fool me once,
25:33 shame on you,
25:33 fool me twice,
25:34 shame on me.
25:35 And I think people in our,
25:36 in our treatment groups,
25:37 you tell them all this fancy stuff
25:39 about insurance and,
25:41 you know,
25:41 how to,
25:42 how to make complaints and how to do this and that,
25:43 the magic bean seeds or maize seeds in this case,
25:47 and
25:47 they'll go,
25:48 well,
25:48 maybe.
25:49 But if they don't see it working,
25:50 and it's a pretty large jump in their expenditures,
25:53 which it was for these farmers,
25:55 uh,
25:55 even when they were dabbling in it,
25:56 they say,
25:57 well,
25:57 that's enough.
25:57 So,
25:58 the,
25:58 the question is,
25:59 for me,
26:00 smart subsidies
26:01 are subsidies that allow people to learn,
26:03 which is a real public good,
26:04 and it spills over
26:05 really rapidly.
26:07 Um,
26:07 but here,
26:08 in fact,
26:08 what we see,
26:09 it's actually really hard to learn about these
26:12 kinds of technologies in the short term,
26:14 and I think that's something we still need
26:16 as kind of a community need to,
26:18 to grapple with.
26:19 So I think I've pretty much said,
26:20 uh,
26:21 most of this.
26:22 So,
26:22 um,
26:23 um,
26:24 thank you very much for your attention.
26:25 Excellent.
26:31 Thank you very much,
26:31 Michael.
26:32 Over to you,
26:33 Matt.
26:34 That was a brilliant presentation,
26:35 Michael.
26:35 It was very,
26:36 really well done.
26:38 So I understood most of it.
26:41 I,
26:41 I don't use this fancy PowerPoint stuff,
26:44 so,
26:44 uh,
26:45 tough room.
26:46 So to,
26:47 uh,
26:47 to prepare for my talk,
26:48 I typed into chat GPT,
26:50 uh,
26:50 give me 1000 words on adapting to climate change,
26:54 and I'd like to read you what the computer said.
26:58 Folks,
26:58 in economics,
26:59 in climate economics today,
27:00 as you know,
27:01 there's two major issues.
27:03 How are we gonna reduce,
27:04 uh,
27:05 the world's carbon footprint at a time when the world economy has to grow?
27:09 We need economic growth to,
27:10 to help all of us.
27:12 How will we decarbonize?
27:14 And at the same time there's a second question of as Mother Nature punches us harder,
27:19 how do we configure the rules of the game and incentives such that we can take a punch,
27:24 and I don't mean evolution,
27:26 uh,
27:26 we're not gonna talk about that in my remaining 9 minutes.
27:29 What I wanna talk about is how do we.
27:31 Set up our markets and the interplay between government and the private sector
27:36 such that we build up resilience.
27:37 I love Michael's presentation.
27:39 How do we build up human capital?
27:41 How do we unleash entrepreneurship and innovation
27:44 such that we get better and better at adapting
27:47 to the scary challenges we face,
27:49 folks,
27:50 for better or worse,
27:51 I'm an eternal optimist.
27:52 My mother says,
27:53 Matthew,
27:53 no magical thinking,
27:55 but I am trained at Chicago.
27:57 Tough room.
27:59 And
28:00 while
28:01 I am a believer in free markets in helping us to adapt,
28:05 I recognize the central role that government can
28:08 and often does play in unleashing the free market.
28:12 But I will also be talking today
28:14 about cases,
28:15 Sam Peltzman cases where government intervention has muffled the free market.
28:20 And want to discuss a research agenda that I'm
28:23 part of and where I have many open questions,
28:26 uh,
28:26 since I'm gonna do no theory,
28:28 since I'm gonna show you no empirics and since you're checking your cell phones,
28:32 uh,
28:32 I want you to know that what I do on a day to day basis is,
28:36 uh,
28:37 as an empiricist
28:38 and as a soft theorist I am always looking
28:41 for empirical benchmarks of are we doing better.
28:44 At adapting to climate change,
28:45 for example,
28:46 we released an NPR paper.
28:48 Sean McLal was one of our co-authors and great to work with him
28:51 with repeated floods that have taken place over the last decade,
28:55 our lights at night dimming less in the aftermath of a flood and recovering faster.
29:00 That's an example of the types of benchmarking I do
29:04 and I'm always interested in for different measures for people.
29:08 For places for farms when the same punch occurs,
29:12 a flood,
29:12 a heat wave,
29:13 how much economic damage occurs coming back to Michael's talk,
29:17 how much would we be willing to pay
29:19 to not face the shock
29:21 and evidence of adaptation is if we're actually willing to pay less over time
29:26 because we're getting better at adapting to the shock
29:28 and so that's sort of the structural benchmark I'm looking for.
29:32 Folks,
29:33 when I was a Gary Becker student and I was not the star of the class,
29:36 I was in Ed Glazer's shadow,
29:38 um,
29:39 what we often were talking about was the household production theory.
29:43 People want to be safe.
29:45 They want to raise happy,
29:47 comfortable children
29:48 and,
29:48 and,
29:49 and productive children.
29:50 Let's not forget that,
29:51 uh,
29:51 in a world where we face increased risks.
29:54 How do we achieve these goals?
29:56 How,
29:56 I'm gonna read from my script.
29:58 How do we remain safe and comfortable in an increasingly risky world,
30:02 folks,
30:02 I wanna start with the boring air conditioner,
30:05 and,
30:05 and can I get a boo?
30:06 I'm looking at intermitt.
30:07 The,
30:07 the,
30:08 the beauty of the air conditioner
30:10 for
30:11 to start a discussion about adaptation is there's a 1 to 1 function.
30:15 If it's extremely hot.
30:17 Side,
30:18 people
30:19 who can afford an air conditioner can buy it and operate it,
30:22 folks.
30:23 There was a paper,
30:23 oh,
30:24 my teleprompter just went out.
30:26 Uh,
30:26 there was a paper in the Journal of
30:28 Political Economy six years ago by Michael Greenstone,
30:31 Karen Clay,
30:31 Joe Shapiro,
30:32 and a few more co-authors
30:34 called The Remarkable,
30:36 it had the word remarkable in the title,
30:37 and here's what it documented.
30:39 In Iowa in the 1930s when it was over 95 degrees for a few days in the summer,
30:45 many people died.
30:46 We didn't have air conditioning and Americans suffered circa 1930.
30:50 So these guys ran regressions at the state level of how many
30:54 people died per capita as a function of how many hot days,
30:57 and they ran this regression every decade
31:00 and showed that over the 20th century,
31:03 fewer and fewer.
31:04 Americans are dying in heat waves,
31:06 and we're not evolving.
31:07 They attribute the success to air conditioners becoming cheaper,
31:11 of higher quality,
31:12 and Americans growing richer to afford these.
31:15 We anticipated a challenge.
31:17 We demanded the product and we installed it.
31:19 Even my mother-in-law in Berkeley is now considering air conditioning.
31:24 And let's be careful,
31:24 air conditioning is not free.
31:26 There's a I have an alternative hypothesis.
31:29 So I think a lot of Indians are moving to Iowa and for us 95 degree days are fantastic.
31:34 So this,
31:35 that was not part of their model.
31:36 This was a closed economy and that,
31:38 but,
31:38 but this is why intermitt's the chief and I have 6 more minutes.
31:41 The,
31:41 um,
31:42 it
31:43 folks,
31:44 but the air conditioner is a convenient example for a free market optimist because
31:49 we have it.
31:50 We recognize that heat causes this challenge,
31:52 and we have this offsetting product.
31:55 Um,
31:56 uh,
31:56 here,
31:57 let me say something productive.
31:58 In my city of Los Angeles,
32:00 there is the recognition that there's many poor people,
32:03 and I have a paper in the Journal of
32:05 Public Economics of the challenges that poor people face
32:08 in the tougher neighborhoods in Los Angeles where
32:10 they live in older housing without air conditioning.
32:13 Karen Bass and Eric Garcetti.
32:15 Have introduced cooling centers and have worked with Uber and other ride sharing
32:19 companies to bring disadvantaged people to cooling centers on the hottest days.
32:24 So while LeBron James and Kim Kardashian can crank up their air conditioning,
32:28 tough room,
32:30 uh,
32:30 at the same time as they privately adapt to the climate change challenge,
32:34 um,
32:34 I.
32:35 Support my being taxed to provide these services for
32:38 the poor to bring them into these cooling centers.
32:41 So that's a case where the government's efforts to provide
32:46 adaptive services to the disadvantaged do not
32:49 crowd out the efforts by private citizens.
32:52 But folks,
32:53 I'm gonna argue that that's rare.
32:56 There is often,
32:57 uh,
32:58 uh,
32:58 there are often Peltzman effects involved,
33:01 and let me give a classic one.
33:03 Carolyn Kowski,
33:04 who
33:04 would be a great panel member today but could not join us,
33:07 she wrote a tremendous paper 20 years ago
33:10 where she said,
33:11 uh,
33:12 a crowding out hypothesis,
33:13 and here's the Peltzman part.
33:15 If there's a city like Miami that floods,
33:18 if that city builds a seawall.
33:21 That can actually be an attractive nuisance as more people move
33:24 to the area because they perceive it to be safe.
33:27 So folks,
33:27 does everyone see that crowding out hypothesis?
33:30 And so a fundamental idea that I'm eager to work with World Bank scholars about is
33:35 unintended consequences.
33:37 If you build infrastructure at a desirable place
33:40 that helps the incumbents,
33:42 but does it make the place even more desirable to move there?
33:45 And if this infrastructure is
33:47 unreliable,
33:48 if it depreciates over time and isn't.
33:50 Maintained you could actually run into an engineering disaster of luring people
33:55 if they overestimate the safety
33:58 uh there can be unintended consequences of crowding out private self-protection.
34:02 So
34:03 one theme I want to stress today is to
34:05 think about when are government efforts a complement of private
34:09 efforts to improve your safety
34:11 in the face of climate change versus when do government efforts
34:15 unintentionally crowd out private
34:18 adaptation efforts.
34:19 And so that is one
34:21 big theme I want to convey today.
34:23 I think I want to convey two more points.
34:26 I want to take a look at my notes.
34:30 Two more segues.
34:32 There's a very interesting literature out of
34:35 Stanford University on government officials' incentives,
34:39 and I think the following merits much more research.
34:42 So again,
34:43 what this,
34:43 uh,
34:44 non-lucid talk it was supposed to be about
34:46 is about how do we configure public and private incentives such that they're
34:51 dovetailed to help us to increase our resilience as we face greater risk.
34:56 Neil Mahorda of Stanford has written several papers
34:59 that in democracies elected officials are rewarded after a
35:04 disaster occurs for bringing back the federal bacon
35:07 and international bacon to help an area rebuild.
35:11 So after Hurricane Katrina,
35:12 after Hurricane Sandy,
35:14 that the FEMA leader,
35:15 the president shows up and shakes the hands with the mayor of the injured place.
35:20 To me this is a very important issue if this holds in the developing world.
35:24 I want us to live in an economy of Gary Becker's economy
35:28 where rational agents invest in self protection before a shock
35:32 to reduce the risk versus taking the gamble.
35:36 Of course it depends on the specifics,
35:38 but,
35:38 but we currently in democracy elected.
35:41 Officials do not appear to be rewarded for prudent investments that
35:45 reduce low probability events to be even less lower probability,
35:50 if that's proper English.
35:51 Instead,
35:51 elected officials are more rewarded for when a disaster
35:56 occurs for bringing back more pork to their region,
36:00 and that strikes me as very bad incentives.
36:03 The takeaway message that I want you to think
36:06 about is I want government to provide public goods,
36:09 to adapt to climate change.
36:11 We need to know where is higher ground.
36:13 We need to know what parts of each nation face less objective flood risk,
36:17 face lower fire risk,
36:19 face lower PM 2.
36:20 5 spike risk
36:22 we need to up zone and change our land use regulations to build up in those areas
36:28 so that poor people and middle class people
36:30 can live there to avoid the climate gentrification hypothesis
36:34 that posits that only Elon Musk will live on higher ground and I.
36:38 Will try to grab his toe,
36:39 uh,
36:40 as the seas rise.
36:41 I wanna live in a world where we,
36:43 uh,
36:43 climate science identifies
36:45 where is higher ground,
36:47 and we have again land use regulation such that we substitute
36:51 capital for land and build up in relatively safer areas.
36:55 In my California
36:57 we face greater fire risk,
36:59 and we,
36:59 uh,
37:00 the officials are not allowing insurance prices to rise in risky areas.
37:05 That sends bad signals.
37:07 Uh,
37:07 ensure regulatory agencies are not passing on insurance costs
37:12 to those in the fire zone.
37:14 I understand that there's incumbents in these fire zones,
37:16 and we could talk about that,
37:17 but I want government officials.
37:20 I encourage them,
37:21 we'll be better able to adapt to climate change if price signals,
37:25 uh,
37:25 send
37:26 that we not muffle the adaptation signals.
37:30 For insurance,
37:31 for electricity,
37:32 in places where there is drought,
37:33 water prices should rise.
37:35 Economists are smart enough to think about
37:37 nonlinear prices such that we protect the poor
37:40 while simultaneously having other interest groups face market
37:44 the correct price signal for these resources.
37:48 So the point of this wacky talk is that we need PowerPoint,
37:52 but we also need market price signals
37:56 to unleash
37:58 capitalism.
37:59 I jumped in my organization of my slides.
38:02 There's tremendous optimism right now that capitalism
38:05 is going to fuel the green economy.
38:07 We keep hearing about Tesla.
38:08 We keep hearing about green power.
38:10 There isn't equal optimism out there about adaptation progress.
38:15 If water prices rise in places where water is scarce,
38:19 if.
38:19 Electricity prices rise in in places where we have trouble generating power.
38:24 This unleashes entrepreneurship,
38:26 and these ideas are public goods.
38:28 So Paul Romer's ideas about ideas as public goods,
38:33 uh,
38:33 will play a key role going forward in helping
38:36 the developing world as the US and other.
38:41 Research nations do the basic research to
38:44 give us the adaptation innovations we need
38:47 and intermitt,
38:47 I'm gonna stop there.
38:52 That.
38:54 That was a completely different presentation from the first one.
38:57 It was fantastic too.
38:58 Excellent,
38:59 excellent.
39:00 So now you have
39:01 the 3rd presentation by Eko Swehe.
39:04 Uh,
39:04 please go ahead.
39:05 I don't know how you're gonna top those two.
39:06 I don't know either.
39:09 Uh,
39:09 no,
39:09 it's really,
39:10 it's really fascinating to be here and to listen to the presentations,
39:13 um,
39:13 because they touch on so many
39:15 points I think impact the work that I do.
39:17 I'm not a researcher,
39:18 but I'm very much
39:20 involved in the messy world of implementation and precisely,
39:23 as was reflected,
39:25 looking at insurance in the context of resilience and adaptation,
39:27 which is,
39:28 uh,
39:29 uh,
39:29 not a conversation that is often top of mind.
39:31 Uh,
39:31 so I've been asked to speak.
39:32 On mobilizing the private sector to reduce the financial protection gap
39:37 um of vulnerable countries and uh that's really my domain.
39:40 It's really a focus on,
39:41 on developing countries,
39:42 emerging markets,
39:43 um,
39:43 and the protection gap being the difference between total losses,
39:47 um,
39:47 and insured losses.
39:48 Um,
39:49 so maybe it's useful to give a little bit of context given,
39:52 you know,
39:52 what we're,
39:53 uh,
39:54 where we work as the insurance development forum.
39:56 Uh,
39:57 the fact that,
39:57 um,
39:58 the estimated figure lost to extreme natural disaster
40:01 events globally over the past
40:03 40 years,
40:04 uh,
40:04 was 4 is approximately 4 trillion,
40:06 and that's a really conservative estimate,
40:09 um,
40:09 and only 2.9 trillion of which,
40:11 uh,
40:11 and
40:11 US 2.9 trillion of which was uninsured.
40:14 Uh,
40:14 this also sits against a background where only 1%
40:18 of those natural.
40:19 Disaster losses between 1980 and 2004 were insured
40:23 um in developing countries.
40:25 So that's quite significant when you think about the impact
40:28 of these events on these countries given the absence of insurance markets.
40:32 At the same time,
40:33 there's also extensive research that demonstrates
40:35 a 1% increase in insurance penetration
40:38 results in a 22% reduction in taxpayer disaster recovery burden.
40:42 So that's quite,
40:43 I think,
40:43 um,
40:44 a lot of figures that we can,
40:45 we can kind of,
40:46 uh,
40:46 think about.
40:47 Uh,
40:48 when we look at some of the figures coming out from 2022,
40:50 and this is taken from Aeon,
40:52 um,
40:52 economic losses from natural disasters were estimated at 313 billion
40:56 in 2022.
40:58 Approximately 42% of losses were covered by
41:01 insurance.
41:02 Again,
41:02 some significant figures.
41:03 I think that should startle us.
41:05 An important point to note is also that those who actually need protection
41:09 are often uninsured,
41:11 right?
41:11 When you look at this map,
41:13 uh,
41:13 there's a lot of work that insurance companies do in terms of tracking,
41:16 uh,
41:16 what this looks like,
41:17 and I think that this is an important
41:19 point
41:20 to note.
41:21 And
41:21 it's also similar when you look at developed markets.
41:24 Poor,
41:25 vulnerable communities often,
41:27 often also don't have access
41:29 to insurance coverage.
41:31 Again,
41:32 stepping back and thinking about developing countries,
41:36 they also exist in a landscape where the humanitarian funding gap is very real.
41:41 In 2022,
41:42 we saw funding requirements spike up to 51.5 billion.
41:47 And
41:47 the system only received 28.7 billion.
41:50 So that's a significant gap in terms of
41:53 an absence of protection in terms of insurance markets,
41:56 but also an absence in terms of financing from the humanitarian
42:00 community that would often step in to support low-income households.
42:04 And that means there's a real mismatch between the commitment to protect
42:08 and the funding model and the funding architecture that we
42:10 have as a global community in terms of disaster protection.
42:14 That speaks to poor preparedness,
42:16 planning,
42:17 underinvestment in data and systems for for response,
42:21 underinvestments in adaptation and risk reduction,
42:23 which,
42:24 which you had spoken about,
42:25 slow politicized tactical decision-making processes that are,
42:29 that are not rooted in the science
42:31 or the existence of systems,
42:33 fragmented,
42:34 poorly coordinated response
42:36 and just limited learning over the decades that we've been,
42:39 been going through this.
42:40 Um,
42:41 an important fact.
42:41 That we should also note is that when we look at COVID-19,
42:45 when it hits,
42:46 only 2%
42:47 of global funding was planned and pre-arranged in advance.
42:52 That's quite startling when you think about
42:54 if we just spent a little bit of time to think about putting some resources,
42:59 making resources available before shocks occur,
43:02 um,
43:02 we could save significantly,
43:05 but the system is currently simply does not
43:08 do that.
43:09 So what is the work of the IDF?
43:11 Um,
43:12 we're constructed as a public-private partnership,
43:15 um,
43:15 and it's really about how do we bring insurance
43:18 and related risk management capabilities,
43:21 and I emphasize this point,
43:23 related risk management capabilities
43:25 to address this financial protection gap.
43:28 Uh,
43:28 it's an industry-led initiative,
43:30 but it's also co-chaired with the World Bank as well as the United Nations
43:34 and
43:35 for us this is very much.
43:37 Focused on the reality that we are seeing,
43:39 which is that all over the world,
43:40 both developed and developing countries,
43:43 we are seeing a widening protection gap.
43:45 There are skills,
43:46 tools,
43:47 practices,
43:48 disciplines that we can employ
43:50 to do better,
43:51 but we simply are not doing it.
43:52 And so our mandate is to really try to explore
43:55 what are the barriers that we see as an institution,
43:57 but also what are the opportunities
43:59 to break some of those barriers through both public and private.
44:03 Uh,
44:03 collaboration.
44:04 Obviously,
44:04 the protection gap is a,
44:05 is a threat to the sustainable development goals.
44:08 That's obvious.
44:09 But for us as an institution,
44:11 we've essentially decided to focus on five areas where
44:13 we believe there are critical contributions that can be made
44:17 and where I think there's also a tremendous amount of opportunity for learning,
44:21 research
44:22 in this space.
44:23 The first is really on risk modeling and we often hear in developing,
44:27 developing countries.
44:28 Data,
44:29 there's data available,
44:30 etc.
44:31 Sometimes,
44:31 yes,
44:32 the data might be available,
44:33 but what's the quality of the data?
44:34 What's the accessibility of the data?
44:36 How is it being used
44:38 to frame discussions with ministries of
44:40 finance around the critical decisions and trade-offs
44:43 that they have to make?
44:44 Very little of that actually takes place.
44:47 So for us within the IDF,
44:49 there are two components to this when we assess the situation,
44:51 which is there is a structural question.
44:54 That we have to ask in terms of making risk analytic data
44:58 more
44:59 prevalent or more available for those communities
45:02 that need it the most.
45:03 And then there is an information gap which we have to close
45:07 when we talk about driving resilience based on an understanding of risk.
45:11 So in terms of the architecture of risk information.
45:15 One of the things that we are really focused on is
45:17 how do we support open risk modeling frameworks and interoperability standards.
45:22 And this is really to facilitate facilitate
45:24 better sharing and cross-sharing of data,
45:27 not only
45:28 between insurance companies,
45:30 but also with the public sector.
45:31 And one of the,
45:32 I think,
45:32 really positive developments,
45:34 um,
45:34 was work that we've been doing.
45:35 With the World Bank on converting the World Bank's
45:38 risk library into a format which could also be
45:41 used by um some of the platforms that we
45:43 are investing in or the industry is investing in,
45:45 in this space.
45:47 The second component is,
45:48 OK,
45:48 if we tackle the architecture element,
45:51 there still remains the substantive gap that exists in terms of risk information.
45:57 And so one of the things that we've done is to partner with the V20
46:01 group of countries,
46:01 which are 58 of the most vulnerable countries in the world,
46:04 to really establish a system where we can work with
46:06 them to try to tackle some of these questions.
46:09 And again,
46:09 part of this,
46:10 and I won't go into it because of the time,
46:12 is really trying to.
46:13 Find
46:14 institutional structures that can allow for the
46:16 private sector as well as the public sector
46:19 to put in resources towards,
46:21 towards tackling what I think is a,
46:23 is an important challenge and also basically
46:25 the development of a global public good if you can consider it in that way.
46:29 Again,
46:30 I've spoken to the,
46:31 the elements in terms of architecture but also in terms of substantive.
46:35 But beyond that,
46:35 I think one of the issues that is really important is that
46:38 insurers for a very long time,
46:40 we often hear them talk about the protection gap
46:42 and we hear this talked about all the time,
46:44 right?
46:45 But at the end of the day,
46:45 there's a real need for investment in what
46:48 solutions need to look like,
46:51 given what we are seeing with climate change.
46:54 The industry,
46:55 I mean,
46:55 for the past 300 years,
46:57 they are very standard products,
46:58 right?
46:58 But given what we are seeing with the kinds of risks,
47:00 there's a need to rethink
47:02 what are some of the new types of insurance,
47:04 risk sharing,
47:05 risk pooling
47:06 mechanisms
47:07 that are required to take us into the next 300 years.
47:11 So a critical part of that work is how can we create again
47:14 the necessary architecture that will allow for an
47:17 exchange between both the public and private sector
47:20 in designing solutions.
47:22 They be at the sovereign sub-sovereign level
47:24 or also importantly with humanitarian actors who are especially relevant in,
47:28 uh,
47:28 developing countries.
47:30 So a critical part of our work is really
47:31 focused on working with sovereigns and sub-sovereigns on this,
47:35 but this is a complex task,
47:37 and at the center of this is
47:40 public-private partnerships.
47:41 And again,
47:41 I would say the messy reality that is involved,
47:44 and I see you in the mids nodding your head,
47:46 uh,
47:46 of of implementing PPPs,
47:49 uh,
47:49 right,
47:49 um.
47:50 If we talk about this case in in Ghana where government is interested in flood risk,
47:54 there are different types of risk.
47:55 You have large riverline risks,
47:58 um,
47:58 floods,
47:58 but you also have urban floods,
48:00 right?
48:00 And
48:01 those can also be induced by man-made behavior.
48:03 How do you translate that into a risk that's insurable,
48:07 right?
48:07 That's a complex task in terms of,
48:09 do you actually have information on historical loss data,
48:13 right?
48:13 How do,
48:14 how does that
48:15 contend if we are thinking about climate change and future.
48:18 Risks,
48:18 right?
48:19 Are we effectively mapping what that is?
48:22 Um,
48:22 many governments have outdated public asset registries,
48:25 right?
48:26 There are issues in terms of challenges with waste management,
48:28 blocked drains,
48:29 etc.
48:29 So where does the insurance fit in
48:32 into this discussion around risk,
48:33 resilience,
48:34 and adaptation?
48:35 What are the boundaries between investments in risk reduction
48:38 and what that risk transfer needs to look like?
48:40 Those are some of the challenges that we have to
48:43 counter when we get to a phase around implementation.
48:46 But again,
48:47 what's the role of public-private partnerships?
48:50 One of the things that we have seen is that we need
48:52 or the industry needs to work with the public sector hand in hand
48:56 to really address some of these challenges,
48:58 and in 2019 we launched a program with UNDP
49:01 with the support of the German government
49:03 to look at how we can support or work with about 20 countries
49:06 towards addressing some of their risk financing,
49:09 uh,
49:09 challenges.
49:10 And as part of that,
49:11 it also involved again thinking through how
49:13 do you align incentives between private sector
49:16 and public sector,
49:17 who pays for what,
49:18 right?
49:18 What are the outcomes.
49:19 Outcomes that we expect.
49:20 How do you even get
49:22 what are traditional insurance companies that
49:24 compete against each other to collaborate,
49:26 um,
49:27 in such a context and so this is,
49:29 I think,
49:29 represents quite an interesting model and I can go
49:31 into a little bit more depth around its structure,
49:34 um,
49:34 and what we have,
49:35 we have delivered with that program in a short space of time.
49:38 But essentially what we are seeing is that there
49:40 is a possibility to work faster and better,
49:43 right,
49:44 through
49:45 public-private partnerships that are designed appropriately.
49:49 They are difficult,
49:50 they are messy.
49:50 There are lots of lessons that we are learning,
49:52 but there is an opportunity,
49:53 I think,
49:53 to really sharpen that,
49:54 that convergence between these communities,
49:57 and we are seeing
49:58 quite a lot of work going on globally
50:01 with governments
50:02 and sub-sovereign entities really looking to,
50:04 um,
50:05 harness,
50:06 uh,
50:06 different types of risk uh financing solutions.
50:09 They range from,
50:09 as was mentioned earlier in the Caribbean with the catastrophe
50:12 risk pool.
50:13 Which was supported by the World Bank
50:15 or in Africa
50:16 with the African risk capacity supported by the
50:18 World Food Program and the African Union,
50:20 or even sovereigns themselves
50:22 who are making these investments.
50:23 If you look at China and,
50:24 uh,
50:25 uh,
50:25 and some of the work that they're doing with agricultural insurance,
50:28 and it takes different forms in different
50:30 countries depending on what the objective is
50:33 that the government or the counterpart of or the partner,
50:36 uh,
50:36 might have.
50:37 But beyond looking at sovereign and humanitarian.
50:40 We also need to look at how are we developing markets
50:43 because you also have a tremendous middle class in many of these countries
50:48 and a significant portion are people who have
50:49 just emerged out of poverty and with one shock
50:53 go right down
50:54 into that into the,
50:56 into a,
50:56 a,
50:57 a real,
50:57 you know,
50:58 difficult circumstances.
50:59 And so one of the things that we are doing
51:01 as the IDF because we've seen and the research shows
51:04 countries where you have
51:05 robust insurance markets,
51:07 capital markets,
51:08 as well as banking,
51:09 uh,
51:09 sectors.
51:10 are able to recover
51:12 faster,
51:13 right,
51:13 from disasters.
51:15 Where one is missing insurance markets,
51:17 you see the consequence of that.
51:19 So,
51:19 a critical part of this effort is how do we work in the development
51:22 of these markets from the bottom up as opposed from the top down.
51:27 And that involves working with regulators.
51:29 How are regulators viewing their role in
51:32 terms of market development versus consumer protection,
51:35 right?
51:35 How are regulators in emerging markets
51:38 dealing with the rapid evolution that we are seeing
51:40 in market develop in development and product development in Western
51:44 developed markets
51:45 compared to what exists in developing countries.
51:48 The gap is huge,
51:49 right?
51:49 So,
51:49 there's a tremendous amount of work that needs
51:52 to go into development of um these markets.
51:55 But importantly,
51:55 as I alluded to,
51:56 there's also the fact that we are talking
51:58 about insurance markets that have to be regulated,
52:00 right?
52:01 So there is work to be done in terms of engagement
52:03 with regulators and we saw recently
52:05 the International Association of Insurance Supervisors
52:08 issue a statement which was quite unusual,
52:10 really encouraging regulators
52:13 to see the protection gap as an issue,
52:15 right?
52:16 And to be more proactive in terms of their policies around,
52:19 uh,
52:19 uh,
52:20 um,
52:20 tackling,
52:21 tackling this.
52:23 And I just referenced some papers which you can,
52:24 you can look up on the IDF website that speak to not only regulatory dimension,
52:28 maybe opportunities in terms of leveraging technology
52:31 to address some of these gaps.
52:33 And then finally,
52:34 one of the things that I think that we are also exploring as the,
52:37 as the IDF is looking at insurers not
52:39 simply as underwriters but also as institutional investors.
52:42 Um.
52:43 It's challenging.
52:45 Most insurers don't invest in
52:48 developing markets.
52:49 Uh,
52:49 there are a variety of reasons,
52:50 capital standards,
52:52 um,
52:52 also constraints around project pipeline,
52:54 etc.
52:55 but we see this as an important space in terms of
52:58 if we are interested in adaptation,
53:00 resilience,
53:01 growth,
53:01 resilience,
53:02 and all these issues,
53:03 we hear
53:04 the conversations that are happening.
53:06 Around the need to mobilize financing,
53:08 then what is the opportunity for insurers to lean into this space?
53:11 There have been some,
53:12 I think,
53:12 innovative work,
53:13 uh,
53:14 done with the IFC on the MCCP program,
53:16 uh,
53:17 through some innovative vehicles that allow for that,
53:19 um,
53:19 but clearly there is a need for more and I think
53:21 an appetite for more also when you think about the long term
53:24 interests of,
53:25 of insurers.
53:26 So
53:27 just to conclude.
53:28 I,
53:28 I think,
53:28 you know,
53:29 this raises some,
53:30 some interesting points for us when we think about growth
53:33 and resilience because for the work that we do,
53:35 it's about
53:36 how do we build effective risk management and protection systems.
53:40 With climate change,
53:41 we're seeing new risks,
53:43 complex interconnected risks.
53:45 We're also seeing a demand in terms of risk management tools and services.
53:50 We're also seeing the need to deepen our
53:53 engagement with governments on improving public financial management,
53:56 and that's challenging because a lot of what we do
53:58 doesn't necessarily reconcile with political cycles,
54:01 as was mentioned earlier,
54:02 right?
54:03 So how do you contend
54:04 with some of the initiatives that require a long-term investment.
54:08 There's also the issue of transition risks and
54:10 opportunity that and the limit the challenges,
54:12 but also opportunities that come
54:13 with that if we step away from the resilience and adaptation,
54:16 uh,
54:17 discussion.
54:18 But it's also about keeping pace with innovation.
54:21 How do we drive and use,
54:22 um,
54:23 uh,
54:24 risk information when we see the rapid evolution
54:26 that is taking place with technology firms,
54:28 how is that filtering into developing countries,
54:31 right?
54:31 And how that information is being absorbed
54:34 into in um
54:35 into informing.
54:36 Action,
54:37 right?
54:37 How are we ensuring appropriate regulation,
54:39 as I mentioned earlier?
54:41 What is the investment in new product development,
54:44 right?
54:45 We can talk about the protection gap and all these different things,
54:47 but are we willing to be innovative and invest and think creatively,
54:51 uh,
54:51 about what these products and feel and learn.
54:54 Right,
54:55 um,
54:55 from what we are doing,
54:56 how are we leveraging technology.
54:58 But central to that is also about fostering collaboration and
55:02 coalitions,
55:03 driving greater scale through knowledge sharing,
55:05 improving systems,
55:07 new structures.
55:08 And focusing on solutions,
55:11 right?
55:11 Learning,
55:12 implementing,
55:13 yes,
55:13 there is some failure,
55:14 but how do we improve a constant process of improvement,
55:17 but also as part of that,
55:19 providing public value.
55:20 It's not just simply about insurance,
55:22 it's about the ecosystem that is created.
55:25 The discipline of it,
55:26 putting resources aside to deal with the potential shock,
55:29 but it's also
55:30 about,
55:31 you know,
55:32 the investment through information that we should be making in fundamental risk
55:36 risk reduction efforts
55:37 that is also part
55:39 when we think
55:40 about growth and resilience of what we need to be doing.
55:42 And I will just close with a little bit of,
55:44 uh,
55:44 maybe a little bit of an anecdote which is
55:46 we're currently having a conversation internally within the IDF
55:49 and amongst many of our members,
55:51 right,
55:51 insurance actors,
55:52 because we are seeing the impacts of these risks.
55:54 We talk about California and State Farm,
55:55 right,
55:56 and the recent
55:57 withdrawal.
55:57 There's a role in terms of
55:59 Sending a signal in terms of price.
56:01 But when you leave vulnerable communities exposed,
56:04 that's a huge reputational risk on every party involved.
56:07 So what is the opportunity for public-private partnerships to figure
56:11 out a way in which we can send those signals
56:13 but also don't leave people stranded.
56:15 So,
56:16 one of the things that we've also been,
56:17 we've been discussing is when we think about
56:20 the 1890s
56:21 when there was,
56:22 we saw the increasing uh electrification,
56:24 engineers,
56:25 industry,
56:26 insurance industry sent out engineers
56:28 to consider,
56:28 should we support this,
56:30 right?
56:31 Or is this a major issue that we,
56:33 we need to,
56:33 you know,
56:34 really pull back in terms of our,
56:36 our investments
56:37 and the engineers,
56:38 I can't remember his name,
56:39 but he was at MIT,
56:40 came back and said
56:41 electrification is inevitable.
56:43 It is the future,
56:44 right?
56:45 What you need to do as insurers is find a way to make it safe to occur,
56:49 right?
56:50 And that was what led to the creation of Underwriters Laboratory,
56:54 which is basically an an entity
56:56 that certifies all electrical equipment
56:59 that you see,
56:59 most electrical equipment that you see around the world.
57:02 So there is a conversation that we are pushing with our members,
57:05 with the industry members,
57:06 but also with the public sector.
57:07 We are faced with a world of increasing uncertainty.
57:11 Huge shocks
57:12 we have to reimagine
57:14 what those solutions need to be for the future and it cannot simply be
57:18 yes with withdrawal in some instances that's appropriate,
57:21 but also trying to think about what does growth and resilience
57:24 look like in the future when we think about climate change
57:26 and what are the opportunities as.
57:29 An institution
57:30 to unlock that and make that happen.
57:32 So
57:33 I'll end there and say that's the work of
57:35 the IDF and hopefully,
57:36 hopefully we can um
57:38 collaborate with others on that.
57:39 Excellent,
57:40 excellent.
57:40 IDF has an excellent general secretary.
57:46 So I attended some of the earlier sessions I,
57:49 I saw a Fast and Furious session.
57:51 This one
57:52 was fast,
57:53 furious,
57:53 and fantastic.
57:54 This,
57:54 this was brilliant.
57:55 This was really,
57:56 really good.
57:57 Uh,
57:58 so now we have,
57:59 we have some time for you to ask questions,
58:01 and I,
58:02 uh,
58:03 uh,
58:03 so I would encourage you to do that.
58:05 Can you,
58:06 uh,
58:07 who wants to go first?
58:08 Or do you want to go home?
58:12 Anybody?
58:13 Yes,
58:13 sir.
58:13 Come on,
58:14 please.
58:15 Just introduce yourself and ask the question.
58:17 Sure.
58:18 Hi,
58:18 uh,
58:19 Rodrigo Garcia from,
58:20 uh,
58:20 Sua Climate Insurance,
58:22 and I,
58:23 I find like the discussion fascinating and it's been mostly about,
58:26 um,
58:27 uh,
58:27 risks transferring,
58:28 right?
58:28 And probably to professors can point,
58:31 uh,
58:31 risk mitigation is something that probably we need the most,
58:34 it could be the most efficient
58:36 in terms of incentives,
58:37 but these two,
58:39 pieces are.
58:40 Uh,
58:40 are filled with market and government failures.
58:43 So,
58:43 uh,
58:44 I,
58:44 I,
58:44 I agree like it's necessary to think through like uh
58:47 economic efficiency,
58:48 but,
58:48 uh,
58:48 I don't see clearly how to think,
58:50 uh,
58:50 like what is a framework to think through these problems
58:53 to begin with.
58:57 Do we have any other questions
58:59 we can ask all at once.
59:01 And then we can,
59:03 and then I also want to thank the organizers at the end.
59:06 So I don't want you,
59:07 I don't want you to leave before then.
59:09 Please
59:09 continue.
59:10 OK,
59:11 uh,
59:11 thanks inermit.
59:12 Uh,
59:12 I'm Tali Kulli.
59:13 I'm with the,
59:13 uh,
59:14 World Bank.
59:14 Uh,
59:15 I have a question
59:16 for Michael.
59:18 Uh,
59:18 Michael,
59:18 so on resilience measurement,
59:20 there's,
59:21 uh,
59:21 obviously there has been an active literature,
59:23 uh,
59:24 particularly over the last 1015 years.
59:26 Um,
59:26 I'm not sure if you've seen
59:28 Chris Barrett's last paper,
59:29 a recent paper on this comparative assessment of
59:32 various resilience approaches where he basically casts doubt
59:36 on existing measures,
59:37 uh,
59:37 and their utility and encourages further research.
59:39 So
59:40 I wanted you to contextualize a little bit your
59:43 measure vis a vis existing measures and in particular.
59:47 Uh,
59:47 what you see the methodological frontier,
59:50 uh,
59:51 for
59:51 the development of that particular measure and how
59:54 longitudinal survey platforms such as the LSMS could actually help propel
59:58 that particular agenda.
59:59 Thank you so much.
1:00:00 Thank you,
1:00:00 Taleb.
1:00:01 Next one.
1:00:02 Next question,
1:00:02 please.
1:00:03 We'll take them all at all at once.
1:00:04 Hi,
1:00:05 I'm Chavi Gine from Deck DCRG uh here at the bank.
1:00:08 So this is also a question for Michael,
1:00:11 uh,
1:00:11 and then for Ecasue,
1:00:13 um,
1:00:14 so.
1:00:15 So,
1:00:16 so I,
1:00:16 I think I missed that,
1:00:17 but,
1:00:17 but basically when you compare the control to the
1:00:19 treated or the controls that had experienced a shock,
1:00:23 so
1:00:24 did you say what happened to the treated that did not experience a shock,
1:00:28 uh,
1:00:28 and so how does that,
1:00:29 so is that the more natural comparison group to the,
1:00:32 to the treated that did experience a shock and then,
1:00:35 so I think you,
1:00:35 I think you mentioned that actually they.
1:00:38 Those that did not experience the shock but were,
1:00:40 you know,
1:00:40 had the seeds and,
1:00:41 and,
1:00:42 and insurance
1:00:43 basically did not,
1:00:44 you know,
1:00:45 behaved as controls,
1:00:46 I believe,
1:00:47 um,
1:00:48 but then,
1:00:48 you know,
1:00:48 this is only after one year so
1:00:51 even,
1:00:51 even if I,
1:00:53 you know,
1:00:53 even if,
1:00:54 if I'm a farmer and experience a shock,
1:00:56 um.
1:00:57 You know,
1:00:58 and I'm more likely to buy,
1:00:59 to buy the bundle,
1:01:01 you know,
1:01:01 what happens in subsequent years,
1:01:02 right?
1:01:02 I mean this is not,
1:01:03 you know,
1:01:04 or,
1:01:04 or,
1:01:05 you know,
1:01:05 do I learn in one period and then,
1:01:07 and then I'm,
1:01:07 you know,
1:01:08 even though
1:01:09 I may not benefit later on,
1:01:10 I still keep on buying this,
1:01:12 right?
1:01:12 And so,
1:01:12 so how,
1:01:13 how should we,
1:01:13 how should we think,
1:01:15 how should we think about this?
1:01:16 and then,
1:01:17 you know,
1:01:17 a related question and so,
1:01:19 so this is again,
1:01:19 these are folks using the same technology,
1:01:22 presumably growing the same kind of crop.
1:01:25 But,
1:01:25 but,
1:01:26 but there's another counterfactual which is,
1:01:28 you know,
1:01:28 are these,
1:01:29 should be,
1:01:29 should be these guys be,
1:01:30 you know,
1:01:30 given the conditions,
1:01:31 given climate change,
1:01:33 should be these guys be growing
1:01:35 whatever it is that they're growing,
1:01:36 or should they just,
1:01:38 you know,
1:01:38 uh,
1:01:38 should,
1:01:38 should
1:01:39 incentives to be provided to.
1:01:41 Either migration,
1:01:42 move away from the farm,
1:01:43 or,
1:01:44 you know,
1:01:44 and,
1:01:44 and,
1:01:45 and,
1:01:45 and you know what,
1:01:46 what would the
1:01:46 what would the resilience gap be if,
1:01:48 if you know with these with these alternative uh kind of policies
1:01:51 and then sorry with with your,
1:01:53 uh,
1:01:54 so you know fascinating uh uh presentation,
1:01:57 um,
1:01:57 but I think you know what what I was after is is what you said at the very end which is.
1:02:03 So I think we're in an environment where there are places,
1:02:06 you know,
1:02:06 coastal Florida,
1:02:07 California,
1:02:08 where actually
1:02:09 private insurance companies are pulling out
1:02:11 now I don't know what's happening in,
1:02:12 in Africa in terms of,
1:02:14 you know,
1:02:14 uh,
1:02:15 uh,
1:02:15 index insurance or so on and so forth where again the prevalence of,
1:02:19 of payouts is increasing and so,
1:02:21 so how,
1:02:21 how is the industry responding,
1:02:23 uh,
1:02:24 uh,
1:02:24 and what are you guys doing about that,
1:02:26 uh,
1:02:26 uh,
1:02:27 um,
1:02:28 yeah sorry,
1:02:28 thank you.
1:02:29 Good,
1:02:29 good,
1:02:29 good,
1:02:29 good,
1:02:30 good question please,
1:02:31 please ask.
1:02:32 Um,
1:02:33 yeah,
1:02:33 hi,
1:02:34 uh,
1:02:34 I'm Claudia from the research department,
1:02:37 and I wanted to pick up on Matt's point about,
1:02:39 uh,
1:02:39 unintended consequences of some of these,
1:02:42 uh,
1:02:43 measures,
1:02:43 for instance,
1:02:44 uh,
1:02:45 the,
1:02:46 uh,
1:02:46 air conditioning which made it,
1:02:48 uh,
1:02:48 more habitable to live in
1:02:50 super high countries.
1:02:52 They also,
1:02:53 uh,
1:02:53 um,
1:02:54 contributed to
1:02:56 the
1:02:56 hole in the ozone layer back.
1:02:58 When I was a kid,
1:03:00 which the Globe
1:03:01 was successful in.
1:03:04 Getting together to ban those chemicals and uh
1:03:08 so I like that as both the unintended consequence for the ozone layer
1:03:13 as well as the world getting organized to address the problem,
1:03:17 but as far as I,
1:03:18 as far as Google tells me they continue to
1:03:21 contribute to greenhouse gasses.
1:03:22 So,
1:03:24 uh,
1:03:24 I just wanted to
1:03:25 pick up further on that point.
1:03:27 Thank you.
1:03:27 Good question.
1:03:30 Hi,
1:03:31 uh,
1:03:31 I'm Alessandro Barratiri from the World Bank,
1:03:33 and a short question for Xuedi.
1:03:35 I'm not an expert of this.
1:03:36 It was,
1:03:36 uh,
1:03:36 very fascinating,
1:03:38 and I was wondering if you could
1:03:39 tell us a little bit more about the history of the
1:03:42 institution,
1:03:42 so the,
1:03:42 the,
1:03:42 the,
1:03:43 the incentives,
1:03:44 and,
1:03:44 uh,
1:03:45 that have led the industry to,
1:03:46 to,
1:03:47 to provide this,
1:03:48 uh,
1:03:48 uh,
1:03:49 proposal of PPP,
1:03:50 and so the scalability also,
1:03:51 now you mentioned 5 billion clearly is,
1:03:53 is a small drop in the,
1:03:54 in the need,
1:03:55 and so how we go forward with that.
1:03:58 Excellent set of questions.
1:04:00 So why don't we go first Ecowe,
1:04:02 then Matt,
1:04:02 then,
1:04:03 and then Michael.
1:04:04 Yeah,
1:04:05 yeah,
1:04:06 great.
1:04:06 Uh,
1:04:06 so I'll start with the history,
1:04:08 um,
1:04:08 of institutions
1:04:10 of the institution.
1:04:11 Actually it came from quite a lot of
1:04:13 conversation that was happening within the UN system
1:04:15 for many years around shocks that were happening in developing countries
1:04:19 and just the concern in.
1:04:20 In terms of how do they work,
1:04:22 can they work with the private sector
1:04:24 to improve the financing for response.
1:04:27 So it was actually through that that there was a concerted
1:04:29 effort to see how can we create an institution where the industry
1:04:33 could work as a collective with the public sector,
1:04:36 uh,
1:04:36 because I mean there were so many institutions,
1:04:38 if I can put it that way,
1:04:40 from UNWFP to UNDRR to World Bank.
1:04:43 Institutions
1:04:44 that the feeling was let's create a single entity
1:04:46 where there could be a collective approach to this.
1:04:48 So that was the history and it was formally launched at COP 21
1:04:52 in Paris,
1:04:53 um,
1:04:54 not
1:04:54 because we are only focused on climate change,
1:04:57 but because that was when,
1:04:58 you know,
1:04:58 it was the opportune time for that.
1:05:01 So
1:05:01 again,
1:05:01 I hope that clarifies in terms of,
1:05:04 of,
1:05:04 of the history of the institution.
1:05:06 It's hugely important because I think that's the first
1:05:08 good thing that's come out of the UN.
1:05:12 I,
1:05:13 I did not say that.
1:05:15 Uh,
1:05:15 I think,
1:05:16 I think it's,
1:05:17 uh,
1:05:18 it's,
1:05:18 it's a good thing that has come out in terms of collective action,
1:05:23 um,
1:05:23 and I think,
1:05:24 you know,
1:05:24 there are lots of opportunities to
1:05:26 learn from that.
1:05:31 Um,
1:05:31 so,
1:05:32 in terms of the second question,
1:05:33 which was,
1:05:34 um,
1:05:34 the industry's response when you took,
1:05:37 when you think about climate change,
1:05:38 I will say,
1:05:38 OK,
1:05:38 we tackle this on two fronts.
1:05:41 With a focus on emerging markets.
1:05:44 We've focused a lot in terms of improving
1:05:46 the engagement at a global policy level around putting this issue on the map.
1:05:52 You would be surprised that this conversation that
1:05:55 we're having here in terms of the protection
1:05:57 gap is not one that is prevalent or often understood or even seen in many contexts.
1:06:03 So that for us is a critical contribution that we
1:06:06 have made or we are making to this space because
1:06:09 without that acknowledgement,
1:06:10 there's very little investment that will actually go into tackling the problem.
1:06:15 The second component is
1:06:16 what do you do in terms of the practicalities around this,
1:06:19 and I,
1:06:20 and as I said earlier,
1:06:21 we are talking about developing countries where you have
1:06:24 between 1 to 3% in terms of insurance penetration rates.
1:06:28 So part basically non-existent,
1:06:30 right?
1:06:31 And so there is some fundamental work that needs to be done in terms
1:06:34 of thinking about how do we actually develop those markets in the first place.
1:06:38 But there's an added layer of complexity indeed,
1:06:41 as you alluded to in terms of climate change and what this introduces.
1:06:44 So a lot of what we do is really looking at different types of
1:06:47 risk pooling mechanisms because that is one way in which you can manage that.
1:06:51 Uh,
1:06:51 the initiative that we see in the Caribbean,
1:06:53 in Africa,
1:06:54 in Asia,
1:06:54 and the Pacific are really aimed at bringing countries
1:06:57 together to pool their risk into a single portfolio
1:07:00 so that they're able to access the market.
1:07:02 At a much better rate.
1:07:04 The second is that we are also seeing innovation
1:07:06 in terms of the products that are being developed,
1:07:08 for example,
1:07:08 in terms of parametric insurance products,
1:07:11 and these are quite different from the traditional indemnity
1:07:13 products where you usually have to have a loss,
1:07:15 you go out into the field,
1:07:17 you do the assessments,
1:07:18 then you pay out a couple of months after the loss adjusters
1:07:21 have determined what that impact is,
1:07:23 right?
1:07:23 And that drives up the cost.
1:07:25 But at the same time,
1:07:26 Parametric insurance also is data intensive and so
1:07:29 there is a need for investment in that data
1:07:31 to actually make these products available and viable.
1:07:34 But then there's also a conversation in
1:07:36 terms of fundamental investments in risk reduction.
1:07:40 In many of the countries that we are engaged in,
1:07:42 there is very weak
1:07:44 social protection systems,
1:07:46 very limited investments in resilience and adaptation.
1:07:50 And so insurance can be valuable,
1:07:51 but it has its limitations,
1:07:53 and that's also part of what we need to figure out.
1:07:55 And its value comes in terms of the extremes,
1:07:57 but also through leveraging risk pooling,
1:07:59 um,
1:08:00 and risk sharing mechanisms.
1:08:01 So I hope I've addressed that
1:08:03 question that was raised very nicely done.
1:08:05 And
1:08:06 I,
1:08:06 I
1:08:07 I it's over to you,
1:08:08 Matt.
1:08:08 I know,
1:08:09 I know that you won't answer the question
1:08:10 and you'd like to talk about something else,
1:08:12 but that's fine.
1:08:13 You can.
1:08:15 So.
1:08:17 So on,
1:08:18 on resilience,
1:08:20 uh,
1:08:20 the world is urbanizing and I think that urbanization
1:08:23 and within nations the competition between systems of cities
1:08:27 is gonna play a key role
1:08:29 to allow us to discover there's so much that we know that we don't know of how
1:08:34 we're going to adapt and this competition.
1:08:36 Between cities
1:08:37 within nations to attract migrants
1:08:40 and within a system of cities this competition is going to teach us so much.
1:08:44 Uh,
1:08:44 Claudia asked an excellent question of what are the carbon footprint implications
1:08:49 as all of us adapt by cranking up the air conditioning.
1:08:52 I think we're nicely cooled here today.
1:08:55 And,
1:08:55 uh,
1:08:56 Paul Gertler,
1:08:57 who I think had a great career at the World Bank for a while,
1:08:59 and Lucas Davis had a great 2015 Proceedings of National Academy Sciences paper
1:09:04 exploring this topic in Mexico,
1:09:07 studying the Grilllius adoption curve as people go from being poor to
1:09:10 middle class of how many more people in Mexico own an air conditioner
1:09:15 and,
1:09:15 and what are the resulting greenhouse gas emissions.
1:09:18 So an unintended consequence of adapting is to exacerbate the externality,
1:09:23 especially as you've scale this up all over the world.
1:09:24 World,
1:09:26 I'm optimistic
1:09:27 that President Biden's
1:09:29 inflation Reduction Act is gonna give us a set
1:09:31 of green blueprints that will diffuse around the world.
1:09:35 Uh,
1:09:35 it's gonna be slow.
1:09:36 Sham McLal and I have written about this,
1:09:39 but I do think,
1:09:40 and I'm a,
1:09:40 I'm a fan of the president's subsidy program
1:09:43 because somebody had to be the guinea pig here
1:09:45 to accelerate our decarbonization.
1:09:48 And of course that's the sweet spot of having the win-win
1:09:51 of cranking up the air conditioning without angering Greta Thunberg.
1:09:55 Well,
1:09:55 it should have been called the Temperature reduction Act,
1:09:57 yep.
1:09:59 Michael,
1:10:01 OK,
1:10:01 um,
1:10:02 just a couple of quick,
1:10:03 uh,
1:10:03 responses to very interesting questions.
1:10:06 First,
1:10:06 Rodrigo,
1:10:06 I just wanna,
1:10:07 as someone who lives in the world of parametric insurance
1:10:11 and something I'm actually talking about tomorrow
1:10:13 morning early in California here in DC
1:10:16 for a World Bank seminar,
1:10:18 uh,
1:10:18 I think one of the big problems is defining quality.
1:10:21 For those kinds of contracts,
1:10:23 so we've been working on the definition of quality standards
1:10:26 and,
1:10:27 uh,
1:10:27 Matt and I looked at each other
1:10:29 when when the underwriter lab was mentioned
1:10:31 because that's actually been our model for what's missing
1:10:34 in the parametric insurance market and I think that issue
1:10:37 has to be taken really seriously more generally,
1:10:40 uh,
1:10:40 if in fact we're gonna tap into the potential
1:10:43 of parametric insurance.
1:10:44 OK,
1:10:44 very quickly.
1:10:46 Uh,
1:10:46 Tala,
1:10:47 you asked a little bit about the other
1:10:49 resilience literature.
1:10:51 Um,
1:10:52 I think most of that literature,
1:10:54 and,
1:10:54 and I am quite familiar with Chris's
1:10:56 damning critiques,
1:10:58 uh,
1:10:58 on it.
1:10:59 Um,
1:11:00 a lot of it is measurement of resi what's called resilience capacities,
1:11:03 so people haven't actually looked at shocks
1:11:06 in the way Townsend did or think of some of the classic.
1:11:09 They've actually said,
1:11:10 well,
1:11:10 we think we know what makes people resilient.
1:11:12 So let's look at the distribution of those things and then they try to use that
1:11:17 predictively
1:11:18 to look at future changes in welfare but not as a mitigating factor for shocks.
1:11:22 So the paper that Mo and I have actually written,
1:11:26 um,
1:11:26 you know,
1:11:27 tries to say let's just kind of start over again from basics.
1:11:30 And,
1:11:30 and resilience means your ability to recover from a shock.
1:11:34 So you're not gonna know that unless you observe people
1:11:36 with a shock,
1:11:37 and that leads me to,
1:11:38 to lead to your
1:11:39 second,
1:11:40 uh,
1:11:40 question,
1:11:41 which is about the LSMS data.
1:11:43 So I think that is absolutely gonna be a vital source
1:11:47 uh for for looking at this resilience and I think one of the things
1:11:51 that we need to might wanna think about
1:11:53 is when large shocks do happen
1:11:56 in in between LSMS rounds,
1:11:59 you know,
1:11:59 what's the frequency of observation that you need after it?
1:12:02 The dynamic programming fake data that I was using,
1:12:05 you know,
1:12:06 we could observe things as frequently as we wanted.
1:12:09 You know,
1:12:10 interestingly,
1:12:10 with the much with the annual frequency data from that Mozambique,
1:12:13 Tanzania,
1:12:14 we actually see we can get some stuff and
1:12:16 so I think there's for me this is a great idea
1:12:19 is to harness,
1:12:22 uh,
1:12:22 to harness the LSMS data to actually
1:12:24 do this and,
1:12:25 and,
1:12:25 and really start to build up a catalog
1:12:28 uh of information on that,
1:12:30 uh.
1:12:30 Uh,
1:12:31 Chave asked a couple of really,
1:12:33 uh,
1:12:34 uh,
1:12:34 good questions.
1:12:35 I think we were maybe a little confused on,
1:12:37 on,
1:12:38 on what the
1:12:39 treatment group that wasn't shocked
1:12:41 was doing and sort of how to use the behavior of that group.
1:12:44 So
1:12:45 it's a little bit messy to think about because literally what we see in the data,
1:12:49 there is a small,
1:12:51 smallish ex ante response
1:12:54 to being provided these things that we do see.
1:12:57 In both in all people in the treatment group and then some of them got zapped
1:13:01 and some of them didn't
1:13:03 so there's a little bit of an ex ante response
1:13:06 which is in part is the is part of the benefit
1:13:09 of the resilience technology so that's why we compared.
1:13:13 To the control group,
1:13:14 uh,
1:13:14 but then as I,
1:13:15 as I tried to say in the talk,
1:13:17 eventually
1:13:18 the non-shocked
1:13:20 people who received the res they,
1:13:22 they really dra drastically reduced
1:13:25 walked away from the technology and,
1:13:26 and I think that's an important thing to sort of,
1:13:29 uh,
1:13:29 think about.
1:13:30 I think you,
1:13:30 you also,
1:13:31 um.
1:13:33 Uh,
1:13:34 you raised a really important point.
1:13:36 Uh,
1:13:37 which is,
1:13:38 you know,
1:13:38 it's one thing to help someone keep growing maize
1:13:41 in an arid area of Mozambique,
1:13:42 but is that actually what somebody,
1:13:44 is that actually the best thing to be done?
1:13:46 So,
1:13:46 again,
1:13:47 I didn't really emphasize it in the talk,
1:13:48 but we were bundling together.
1:13:51 A
1:13:52 genetic technology which it makes it more possible for some people to continue
1:13:56 to profitably
1:13:57 grow maize in some places,
1:13:59 but clearly there are gonna be other instances
1:14:01 and probably in those parts of Mozambique in the near future
1:14:05 where that's actually no longer the right decision
1:14:08 and here is where I think
1:14:09 there's kind of a chicken and egg problem that goes on we,
1:14:12 we.
1:14:12 Uh,
1:14:13 we tried to do something like this in the Dominican Republic
1:14:16 and it never really got off the ground,
1:14:18 so the idea
1:14:19 was to create an insurance contract
1:14:22 that could be used
1:14:23 to certify loans for lenders
1:14:26 that lenders would make to small scale
1:14:28 livestock dairy farmers
1:14:30 so that they could invest then in in irrigation
1:14:33 and water control facilities.
1:14:36 So the problem was
1:14:37 without insurance,
1:14:38 the banks didn't want to touch these guys.
1:14:41 Uh,
1:14:41 and then what we tried to do is combine the insurance
1:14:45 to unlock credit,
1:14:46 but then the credit
1:14:48 was being provided in,
1:14:49 uh,
1:14:50 again,
1:14:50 it didn't ever work,
1:14:51 but the idea was that the credit.
1:14:52 Would then be linked
1:14:54 to some of these more transformative
1:14:56 investments like if you're a small scale dairy farmer you don't just
1:15:00 depend on natural pasture
1:15:01 but you build barrages and you build small scale irrigation
1:15:04 schemes or you could think about that being tied even to
1:15:07 crop protection but we sort of need to get from where people are,
1:15:11 which often risk prone farmers have no capital access or very
1:15:14 little capital access
1:15:16 and,
1:15:16 and,
1:15:16 and,
1:15:16 and get them to where they can move ahead.
1:15:18 So
1:15:19 I think your point is extremely.
1:15:20 Well taken,
1:15:21 we
1:15:21 this should not be seen as a way to lock people
1:15:24 into something that becomes increasingly nonprofitable,
1:15:27 but,
1:15:28 uh,
1:15:28 I think we need to think then about ways to blend together
1:15:31 financial tools so that people
1:15:33 can actually get the loans they need to make the adaptive kind of investments that
1:15:37 I think we would all agree would make sense.
1:15:39 So again,
1:15:39 thank you everyone.
1:15:41 Do you wanna add something,
1:15:42 please,
1:15:42 please,
1:15:43 uh,
1:15:43 I wanted to add to that I think that that's
1:15:44 a very important point because even the work that we did
1:15:47 with the African Risk Capacity.
1:15:49 Um,
1:15:49 it made very clear what the risks were in
1:15:52 terms of droughts and the fact that in some context
1:15:55 there should be a shift in terms of some of the agricultural practices,
1:15:59 uh,
1:15:59 but that's very difficult because it also converges
1:16:01 with some of the political and cultural,
1:16:03 uh.
1:16:04 Issues within a particular country,
1:16:06 but I think an important contribution is also
1:16:09 from the conversation around risk is bringing transparency
1:16:12 to what exists,
1:16:13 right?
1:16:14 And using that transparency to drive further
1:16:17 conversations around how do you mobilize financing
1:16:20 for such transitions that need to,
1:16:21 to take place.
1:16:22 So I think this,
1:16:23 the World Bank's work on the CCDRs
1:16:25 is quite,
1:16:26 I think,
1:16:26 supportive and helpful in some of that
1:16:28 thinking that we're seeing evolving in countries.
1:16:30 So the work that you're doing is also very
1:16:33 helpful for the CCDRs because Matt will know this
1:16:36 and others will know this that there's a lot of confused thinking about
1:16:39 what's adaptation,
1:16:40 what is resilience,
1:16:41 what is development,
1:16:42 what is adaptability,
1:16:43 and so on.
1:16:44 So I think it will really help us a lot and I know that
1:16:47 Farhat is working on this to actually distill all of your ideas into a form
1:16:51 that even world bankers can understand.
1:16:53 So,
1:16:54 um,
1:16:55 uh,
1:16:55 so I'm going to ask you all to do two rounds of applause.
1:16:59 The first round,
1:17:00 I would like
1:17:01 to give an applause
1:17:02 to this panel,
1:17:03 to Michael,
1:17:04 to Matt,
1:17:05 to Echo Swe,
1:17:06 and for Had for actually,
1:17:08 uh,
1:17:09 for,
1:17:10 uh,
1:17:11 for,
1:17:11 uh,
1:17:12 for,
1:17:12 uh,
1:17:13 thinking through these things and organizing a great session.
1:17:16 So first applause.
1:17:17 Thank you.
1:17:23 The second round I wanted to,
1:17:25 uh,
1:17:27 uh,
1:17:27 I wanted to thank very much the colleagues,
1:17:30 uh,
1:17:31 who actually organized the ABCDE
1:17:33 and I'm sure you all,
1:17:35 I'm sure you all recognize
1:17:37 and appreciate that it's actually much easier for us to attend a hybrid conference,
1:17:42 but if you actually try to organize a hybrid conference,
1:17:44 you know that it's actually much harder to organize one.
1:17:47 Than to organize,
1:17:48 uh,
1:17:49 you know,
1:17:50 uh,
1:17:51 to organize a purely virtual one or a.
1:17:55 A purely virtual one or a purely in-person one,
1:17:58 right?
1:17:59 So I really thank very,
1:18:01 very much all of those who think,
1:18:03 especially 10 people,
1:18:04 and those are
1:18:06 Nona
1:18:07 and her last name is Karaashvili.
1:18:10 The next one is Tatiana Kleinberg.
1:18:12 Then we have Tale Kili,
1:18:15 uh,
1:18:15 we have Paul Christian,
1:18:16 we have Gabby Calderon
1:18:18 Morta,
1:18:19 we have Rula Yazigi,
1:18:21 and we have Anna Bonfield,
1:18:24 Jonathan Litwin.
1:18:25 Maria Saffa,
1:18:27 and of course Art Crey,
1:18:29 OK,
1:18:29 so big,
1:18:30 big applause to all of us.
1:18:40 And thank you all very,
1:18:41 very much for staying till the end.
1:18:43 Thank you so much.
1:18:44 Bye-bye.
- add-style
- lp-body-content