YT YouTube2Text

YouTube2Text → Transcripts → Supplemental lecture: Demographic transition and development

Supplemental lecture: Demographic transition and development — Transcript

by CEPR & VideoVox Economics · 15,164 words · 2,242 segments · language en · Watch on YouTube

Open in transcriber Summarize with AI Download SRT/VTT

Full transcript

0:00Great. Good morning, good afternoon,
0:02good evening to everyone. We're going to
0:04go ahead and get started. I'm Doug Gan,
0:07one of the one of the organizers of this
0:11course on macro key concepts in the
0:13macroeconomics of development. And I'm
0:15really pleased to welcome everyone to
0:17today's lecture, the next in the series
0:21in this course. Um, I want to thank
0:24Michelle Tertilt from the University of
0:26Mannheim who's going to be lecturing
0:28today on demographic transitions and
0:31development, which is a great topic and
0:33one we couldn't have a better speaker
0:35for. Um, let me just do a couple of
0:38brief introductory remarks before we get
0:40going. First, I want to thank as always
0:44the organizers of the STEG or the
0:48funders of the STEG program. So STEG
0:50which is the structural transformation
0:52and economic growth program is a program
0:56operated by CR in London. I want to
0:59thank the CPR staff Nadine Clark um
1:03Kirsty McNeel, Lauren Wearing, Mandy
1:06Chan and others for their help in making
1:10this all work. We also particularly want
1:12to thank the funders of the STG program
1:14who are the United Kingdom's Foreign
1:16Commonwealth and Development Office. Um
1:19we're really grateful to them for
1:20supporting STEG and we appreciate that
1:24they're doing this as part of the UK aid
1:26effort.
1:27The structure of the course as you know
1:29is we've had a series of lectures on a
1:32range of topics in the macroeconomics of
1:34development. Today's lecture should fit
1:37right in. We are really appreciative of
1:39Michelle for taking the time to deliver
1:41the lecture. Um, people are doing this
1:43as you know and as we keep saying really
1:45out of the kindness and generosity of of
1:49their um of their hearts and as a
1:52professional service and we're really
1:54grateful to everyone who's participated.
1:58The ground rules are the same today as
1:59they have been in the past. Those of you
2:01who are in the front row have control
2:04over your microphones and you're welcome
2:06to ask questions. Michelle says you're
2:08welcome to interrupt with questions if
2:11you have things you want to clarify.
2:13What we would ask is that for those of
2:14you in the front row, if you're not
2:16asking a question, if you could please
2:18keep your microphones muted the rest of
2:20the time. And then for those who are not
2:22in the front row, um please do enter
2:25your questions in the Q&A. If you raise
2:28a hand, we can't actually do much about
2:30it, but please just enter your questions
2:32in the Q&A. um the we'll try to address
2:36those. Michelle will introduce the TA
2:39for today's lecture in just a minute,
2:41but Suzanne, who's the TA, and I and Joe
2:45and perhaps others will chip in to try
2:47to answer questions in the Q&A. And then
2:49at the end, Michelle's going to leave a
2:50bit of time and we'll try to come back
2:52to some of those questions as well. So,
2:55on on that note, and without further
2:57ado, let me turn things over to
2:59Michelle. And again, thank you so much
3:01for making the time. and we really
3:03appreciate your willingness to give your
3:05knowledge and expertise and to give the
3:07lecture today.
3:08Well, thank you very much, Doug, for
3:09this nice introduction and hello
3:11everyone. I'm I'm very excited to be
3:13giving this STG lecture on demographic
3:16transition and development. I'm also
3:18going to sneak in some something on
3:20women's rights because it just fits
3:21right in. It is connected. Um but before
3:24getting started, let me briefly
3:26introduce my TA. So, this is Suzanne. um
3:31she's a graduate student here at the
3:33University of Manheim working right at
3:34the intersection of development and
3:36macro. So she's perfect for this course
3:39and she'll um be giving another TA
3:42lecture um with some more details on the
3:44models uh next week. So Suzanne, maybe
3:47you want to say a few sentences so
3:49everyone can see your face.
3:51So thank you and hi everyone. So I'm
3:54very happy to be the TA for Michelle
3:56today and I'm also very happy to see you
4:00at least those who I will see on Tuesday
4:03next week uh so that I can answer
4:05question. Please feel free to send me
4:07your questions if you have some.
4:11Great. Thank you. So then let me get
4:14started by sharing my screen.
4:17Um yes. So I'll be talking about
4:20demographic transitions and development.
4:22So, well, I thought it's good to take a
4:24step back and ask, what is demography?
4:28So, no, I looked it up in the
4:30dictionary. This is from the American
4:32Heritage Dictionary. So, demography is
4:35the study of population size, growth,
4:37and age structure. Well, what determines
4:40population size? Well, clearly
4:42fertility, mortality, but also
4:43immigration. Um, and that that leads to
4:46population change. So that's we want to
4:48look at and think about the connection
4:50between these forces. Um so population
4:53change specifically and um development.
4:57So mostly I will be talking about
5:00fertility. So only one of these
5:02ingredients determining population
5:05change. I will have nothing to say about
5:07immigration today. I think that's also
5:09very interesting topic. But when we
5:11think about demographic transition,
5:13immigration, you know, doesn't really
5:14play much of a role. Mortality on the
5:17other hand does so let me briefly say a
5:20little bit about mortality because it is
5:23important um but also it's not going to
5:25be the main focus of the talk um
5:28yes so I wanted to sort of get that out
5:30of the way first talk a little bit about
5:32mortality because that is also important
5:35uh and then move on mostly to think
5:37about fertility changes over time.
5:41Um so if you think you know one way of
5:45measuring mortality is you know think
5:47about life expectancy the larger life
5:49expectancy the less people die. So here
5:52this is actually from a textbook from
5:54David while um but I just thought it's
5:56you know really nice pictures and um so
5:59I took it um but this shows how life
6:02expectancy basically over the last 250
6:05years increase very dramatically um over
6:09time in a set of developed countries. So
6:12today most developed countries UK, US,
6:16Sweden, Germany will be right up there
6:18are you know somewhere between 75 and 80
6:20years and if you go further back in time
6:24um you know was much lower you know we
6:26can easily you know think think about
6:28being you know below 40. So you know all
6:31the speakers I think in the seminar
6:34series would be dead basically um had we
6:36been born in 1750.
6:39So, so huge increases over time and over
6:43the course of development. If we're
6:45looking at developed countries today, we
6:48see similar pattern but you know they're
6:51behind they haven't most of them haven't
6:53reached anywhere close to you know 80
6:55years yet uh really depends on the
6:58country on the region a lot of African
7:00countries is more like 55 or 60 years um
7:04and you know Chile for example some
7:06Latin American countries is larger um
7:09but also I mean they're catching up
7:12right so what this picture is showing
7:14this is going back not quite as far back
7:15in time but going basically back to the
7:17early 1900s we also see large
7:20improvements in life expectancy
7:24so if I plotted life expectancy
7:27versus GDP in a cross-section of
7:29countries you also see a very strong
7:32positive relationship right in those
7:34countries where life expectancy is large
7:37or in other words mortality is low we
7:39have much higher GDP per capita very
7:42strong positive relationship and notice
7:44This is plotted on a lock scale, right?
7:46So it is a very strong relationship. Um
7:49so you can ask the question what's the
7:51connection to what role does um life
7:54expectancy as part of you know
7:56demography if you want what what role
7:59has that played in development? Um so
8:03right so this basically this connection
8:04which is very strong empirically raises
8:07lots of different um research questions
8:09important research questions right you
8:12can ask why has mortality declined over
8:14time well we can also ask is this even a
8:17question for for economists or maybe
8:19this is just a question for medical
8:21researchers you know is just exogenous
8:24technological progress about medicine
8:26that that led to this or is this also
8:29driven by sort of um you economic forces
8:33um leading to more money being um you
8:37know um devoted to for example medical
8:39research and so forth. Um but then you
8:41can ask the opposite question right does
8:43development lead
8:45no sorry that was the third bullet here
8:47right so is it um that development
8:50itself causes technological progress in
8:53medicine right so which direction is the
8:55causality or is it that the mortality
8:57and we can also think about morbidity
9:00decreases uh leads to higher GDP and
9:03higher growth so lots of interesting and
9:05important research question right and
9:07you can think about a bunch of I think
9:10important econ economic mechanisms why
9:13declines in mortality could lead to
9:15higher GDP. So I just wanted to mention
9:18a few here. Higher incentive to
9:21accumulate more human and physical
9:22capital if you live longer, right?
9:24You'll have more years to reap the
9:26benefits. So you might want to go to
9:27school longer and that can lead to GDP
9:29and growth. Um also if you think about
9:33especially morbidity and not so much
9:35mortality but if people are less sick as
9:38children um so for example there's a lot
9:40of development research on worms or
9:42malaria lots of sort of you know
9:45diseases that affect the well-being of
9:48children. Well, then they have less
9:50ability to acquire human capital. Or put
9:53differently, you know, if if you're if
9:55you're more healthy as a child, you you
9:57know, going to school is more fun and
9:59more productive and hence human capital
10:00can increase. Um,
10:03we can also think about and there is
10:05some research on this that higher
10:07mortality at least in some context can
10:10be sort of a good thing. I mean maybe
10:13good in quotation marks but in the sense
10:15that there will be fewer people sharing
10:17the same resources at least you know the
10:19ones that are left over and that could
10:21possibly lead to at least higher GDP I
10:23don't know higher well-being I don't
10:24want to say that but if you're thinking
10:26about growth as we usually measure it um
10:30so I just wanted to put a few papers up
10:33here that kind of try to answer these
10:35kind of questions I just mentioned um
10:39I'm don't have time obviously to go
10:40through all of them in det detail, but a
10:44few of them that I found particularly
10:45interesting is, you know, some papers um
10:48really argue that mortality
10:51explains a lot of, you know, poverty in
10:54today's world, especially if we're
10:55thinking about African contexts. Um
10:59then you know an interesting paper an
11:01econometrica pip paper by Mosha Hazan um
11:05is so he argues that the increases and
11:07life expectancy that we've seen over
11:10time cannot really be the reason that
11:13people accumulate more human capital and
11:15hence lead to growth. And his logic is
11:17he documents very carefully empirically
11:19that most of the increases in life
11:21expectancy we've seen over time in
11:23today's rich countries are mostly taken
11:26in terms of leisure. If it's all leisure
11:29rather than people working more years
11:31then it's not so clear what's the
11:33increased incentive to acquire more
11:35human capital. Right? if you just keep
11:37working the same years
11:40um and but in the end just kind of
11:42retire much earlier then at least the
11:44logic is not so obvious. Um and then
11:47just very briefly I mean these are kind
11:49of maybe provocative papers but I wanted
11:51to mention this paper by Alvin Young and
11:54another one by um Yohim F and Niko
11:56Vlander. They argue that an increase of
12:00mortality at least in those contexts and
12:03this is one young is about HIV and the
12:06other paper is about the black death in
12:08Europe. In those particular context an
12:11increase in mortality led to higher
12:14living off standards but I think that
12:16must be quite specific to certain kinds
12:18of contexts. Um overall the mod um the
12:22cross-country data shows no no negative
12:25correlation obviously right I mean it's
12:27the opposite
12:28um and then there's also a large um
12:31empirical micro literature documenting
12:34um how declines and mortality and
12:36morbidity actually raise the standard of
12:38living but those are also often very
12:40specific about deworming campaigns um or
12:45you know more breastfeeding I mean very
12:48specific And there's an excellent survey
12:50summarizing that um in the annual review
12:52of economics by Hoy Blegley. So whoever
12:55is interested in that specific aspect, I
12:57recommend that survey. Um
13:00so that's pretty much all I wanted to
13:02say about mortality. Um but I see
13:04already the first 15 minutes have
13:06passed. I have to see I have a lot of
13:07material. Let's see. But feel free to
13:09interrupt and ask questions. I mean I'll
13:12I'll adapt the talk along the way, I
13:14think. Um but yeah so now let's talk
13:17about fertility the other important
13:19aspect determining demography or
13:22determining population size. Um again
13:25let's start with some data. So this is
13:28in a cross-section of countries again
13:30plotting GDP per capita as a measure of
13:33output with the total fertility or
13:36against the total fertility rate both in
13:382010. So pretty recent data. So this is
13:41from a recent handbook chapter that
13:43Matias doka and I wrote but I mean you
13:45can easily construct this just from the
13:47world development indicators. This is no
13:49no difficult plot right. Um so it shows
13:53a very strong negative relationship
13:56between fertility the total fertility
13:58rate and GDP per capita. Right? So it's
14:01the poor countries that still have many
14:03kids even today or let's say 10 years
14:06ago. But if you had even more recent
14:08data, I'm sure it would I'm pretty
14:10confident it would look essentially the
14:11same. And it's the rich countries that
14:15have basically two kids on average, but
14:17plus or minus, but you know, above, if
14:20you look at this above 10,000
14:23GDP per capita in 2010, you know,
14:26basically these all hover around two
14:28plus minus, you know, it could go as low
14:30as 1.3 for some countries or 2.2 two or
14:342.3 for some others but there's not that
14:37much variation anymore but before that
14:39strong negative relationship.
14:44Um so that was cross-country data. We
14:47can also look at the dem what's called
14:49the demographic transition around the
14:51world. Right? That's the topic on the
14:52title of the lecture demographic
14:54transitions. So um another way of
14:58measuring fertility is the birth rates
15:01per thousand people. It's not exactly
15:04the same as the total fertility rate,
15:06but it's also easier in terms of
15:09collecting data. This is sort of easiest
15:11to measure. I mean, you just need
15:13population register. So that way, it's
15:16easier to go back far in time. You just
15:18count how many births were there per
15:20thousand population. It's not such a
15:23good measure at at a point in time in
15:25terms of fertility of a given woman
15:28because this would be influenced by the
15:29age structure of the population as well.
15:32All right. In some sense, more
15:34interesting is birth per you know
15:36fertile woman and that's not what this
15:38is. Right? So if you have lots of old
15:40people for example who cannot have kids
15:42then of course the birth rate will be
15:43lower. Um but if you if you look at long
15:47time trends this is actually very useful
15:49measure. Um just in terms of comparing
15:52countries at a point in time I is maybe
15:54not so useful. So anyways, what I'm
15:56trying to to to show here and argue here
15:58is over time birth rates fell just
16:02dramatically. So this is different
16:04regions. Look at Europe. This is the
16:06right top panel here. By in 1800 we had
16:10about 30 or you know it even even a
16:13little bit more than 30 35 birth per
16:16thousand population and by now that's
16:18declined to about 10. saw I mean huge
16:22decline and most of the decline uh you
16:25know happened around the turn of the
16:2719th century or let's say between you
16:30know 186070
16:33and then you know 1920 or so and then
16:36what you see here that's the so-called
16:38baby boom that's most pronounced in in
16:40the United States so if you look at the
16:42lower panel here the baby boom this
16:45increase again around you know 1950s
16:49plus my you know You see between 1930
16:51and you know 1960 there was very
16:54pronounced in North America but also
16:56other countries had it. Um but again the
16:59focus here now today is more on the
17:02overall decline not on the baby boom
17:04right that's what's called the
17:05demographic transition this huge decline
17:08of fertility over time. Um here's
17:10another way of looking at that for just
17:12from the United States. Um
17:16and here is yet another measure of
17:19fertility.
17:20This is uh children ever born. If you
17:23want to model it from an economic
17:25perspective, I mean this is the closest
17:28analog to what we are thinking about in
17:30a model. This is literally children ever
17:33born per woman, right? That's how you
17:35would think about it in a model. Um, and
17:38so in the United States in the beginning
17:41of the 19th century, women on average
17:44had basically almost six kids or five
17:46and a half. Um, and that declined
17:49dramatically by, you know, the beginning
17:51of the 20th century. We were already
17:53down to just uh, you know, about 2.4 I
17:57believe here. And then again, here you
17:58see the baby boom. But let's forget
18:00about that. I want to focus on this
18:02dramatic decline. Um,
18:05okay. Any questions so far?
18:08Everyone asleep? Everyone listening
18:10still?
18:12Okay.
18:16This is kind of more a side note. Um,
18:19but I wanted to show it because it's a
18:21not as well-known fact. I mean, this
18:24fertility decline over time, you can
18:26find in any textbook. I mean, you can
18:28find millions of versions of it. It's
18:30not, you know, very novel. I mean the
18:33fact itself is very clear. Um this um is
18:37a bit more nuanced. This is not the
18:39decline in the kids people are having.
18:42There's a decline in the number of
18:43people people are living with. Obviously
18:46is related but you know it's it's not
18:49you know there are other margins here.
18:51Um and what I'm showing you here what I
18:53plotted here is um different measures of
18:56household size. So the so the top one
18:58the blue is just literally how many
19:00people were living together at different
19:03points in time. By 1850 the average
19:05household had six people. By 2000 the
19:08average and this is for the United
19:10States again. By 2000 the average
19:12household had only two you know three
19:15and a half people about three I think.
19:18Um a lot of that is the decline in
19:21fertility. I mean, if you were having
19:23fewer kids and obviously you're also
19:25living with fewer kids, um that's the
19:28last the bottom year, right? It's that's
19:30the average number of kids in each
19:31household. And that decline from about
19:342.8 or so here over that same time
19:36horizon to a little less than one. Um
19:40this is on average, right? Many
19:41households don't have any kids in them
19:42or even people who had kids, but then
19:44eventually the kids move out, so they're
19:46no longer in the household. So there's a
19:47little bit different way of thinking
19:49about it. Um, but it's not just about
19:52the kids. If you look at how many other
19:55adults are in a household, I thought
19:57that's for me that's kind of almost the
19:59most interesting in this picture. That
20:01also declined over time. In 1850, the
20:04average household comprised about three
20:06and a half adults. By now, the average
20:09household has more like two adults. Um,
20:12so and not all of these adults, by the
20:16way, are even related. So this is not in
20:19this picture, but we also computed at
20:21some point how many non-family related
20:24adults are in an average household and
20:26that's pretty much zero today. But if
20:29you go back to 1850, it was quite
20:31sizable. I I forgot the exact number
20:33now, but I think it was almost one. Um,
20:35right? So this could be a board or a
20:37lodger or someone helping, you know, a
20:39nanny, someone working with you, a
20:41servant, um, and so forth. So but that's
20:45phenomenon pretty much disappeared.
20:48Um but this is a little bit of a detour.
20:50So let's come back to the fertility. Um
20:54the fertility decline over time. You can
20:56also see that in the cross-section.
20:59So this is basically you know 150 years
21:02of cross-sections all in one picture.
21:05Right? So each of these different colors
21:08is a birth cohort of women. But then for
21:12each bo birth cohort we put people into
21:15um death size by income essentially a
21:18measure of income. I don't want to go
21:20into the details exactly how the income
21:22is constructed. Um if you're interested
21:24I can tell you more later on. Um it's
21:26also in one of my papers but so let's
21:29just think about the x-axis here as a
21:31measure of income. And then we have
21:34children ever born right? A measure of
21:35fertility. And then what you can see at
21:38any point in time it's strictly downward
21:40sloping and or pretty strongly downward
21:43sloping. Um and then over time right so
21:48if you look at consecutive cohorts as
21:50we're moving from 1828 to basically 1958
21:55you know we're just moving to the right
21:56we're all getting richer again this the
21:58United States. um as we're getting
22:01richer kind of it's almost the whole
22:03crest section kind of moves down on this
22:05same curve, right? It almost looks like
22:08with the exception of this here, that's
22:10the baby boom again. So, leaving the
22:12baby boom aside for a second, it looks
22:14like we're just sliding down the curve.
22:17So the point I'm trying to make here is
22:19that the cross-sectional relationship,
22:22the fact that richer people have fewer
22:24kids than poor people seems that that
22:27relationship seems to almost look the
22:28same as in the whole country is getting
22:30richer and we're all having fewer cats.
22:32So there could be a connection basically
22:34between the decline over time and that
22:38at any point in time there's a strong
22:40negative relationship between income and
22:41fertility as well. Um so I think if you
22:44think about models that explain these
22:46facts is kind of interesting to see how
22:48you know these two types of um you know
22:52time series and and cross-section seem
22:54to be quite related.
22:58Okay. So let's now think about um
23:01theories of the demographic transition.
23:02So again let's you know put this the
23:04cross-section aside for now or for the
23:07rest of the talk in fact um and let's
23:10think about what could explain this
23:12decline in fertility over time uh and
23:15how is it related to economic
23:18development so what could be the logic
23:20so
23:22I think historians or demographers
23:24sociologists they often talk about the
23:26importance of birth control birth
23:28control being invented or being used
23:31more later marriages is often very um
23:35you know often mentioned it I mean most
23:39economists have concluded this I mean
23:40it's interesting as a descriptive way of
23:43thinking about it but it's not so useful
23:46thinking about the economic incentives
23:48the economic forces because later
23:50marriage that's a choice right it's not
23:53like somehow some determines marriages
23:56are all later now and because of that
23:58people are having fewer kids um you know
24:00it's part it's part of the mechanism
24:02rather than the explanation.
24:04Um similarly also with birth control I
24:07mean some form of birth control condoms
24:10um basically have existed for a long
24:12time. Um and then so people just started
24:16maybe using them more or used more birth
24:19spacing. There's all sorts of me um um
24:22methods. The pill was invented much
24:24later. So that can also not right that
24:26was kind of in the in the late60s 1960s
24:30for sure that cannot explain the the
24:32demographic transition in the 19th
24:34century um so let's leave that aside
24:38what about infant and child mortality um
24:41right so that's often often mentioned as
24:44well by demographer so I was saying that
24:47infant and child mortality used to be
24:49very high you know if you go far enough
24:51back in history basically 30% 40% of all
24:54kids were dying. I mean, super high
24:56numbers. Um, and so because of that, you
25:01know, it's kind of logical that if a lot
25:03of kids are dying, you want more kids
25:05and if hardly anyone is dying anymore,
25:07you don't need as many kids to guarantee
25:08the same number of surviving kids.
25:11Surely that must have played some role.
25:14But um for example, Matias Dubka shows
25:16in an interesting paper
25:18that basically
25:21I mean you have to make a distinction
25:23between net fertility and total
25:25fertility. So in the data also really
25:29net fertility has declined
25:32and so that is much harder to get out of
25:34a model then right. So basically most
25:38models would predict yeah the total
25:41fertility including those kids that were
25:43dying that makes sense that that
25:46declines over time as mortality falls
25:48but those models would also predict an
25:50increase in the number of surviving
25:52children. Basically, children get
25:53cheaper now, right? Each surviving child
25:56gets cheaper. You don't have to produce
25:58so many extra ones that might be dying.
26:01So then you should see an increase in
26:03surviving fertility. But the but the
26:04opposite happened, right? So even the
26:06net fertility net you know the surviving
26:08fertility actually declines um to to a
26:11large extent. Um so that itself or at
26:14least not alone can explain the
26:17demographic transition. There must be
26:18other factors. Um, what about the
26:22increasing opportunity cost of time? So,
26:25a lot of people talk about that female
26:26labor force participation has increased.
26:29Um, and because of that, it's just too
26:31costly now to have so many kids. Well, I
26:33would argue the timing is just far off.
26:36Female labor force participation has
26:38increased largely mostly in large
26:40numbers in the 20th century. Most of the
26:44fall infertility
26:46uh was in the 19th century. So it's like
26:48a whole century later. Um, of course, if
26:51we're thinking about there was another
26:53decline in fertility after the baby
26:55boom, right? So in the second half of
26:58the 20th century, there was also a fall
27:00in fertility. If we're talking about
27:01that, surely increases in female labor
27:04force participation must have played a
27:05role. Um, I think but if we're talking
27:08about the demographic transition, which
27:10I think more about that earlier time
27:12period. So let me just show the pictures
27:14again. Right? So this was from the
27:16United States. I'm more talking about
27:18you know this whole decline before the
27:21baby boom and that was essentially over
27:23by the beginning of the 20th century. Um
27:26so that that was not a time period in
27:28which women were entering the labor
27:30force in large numbers.
27:34Okay. Then the
27:38um a lot of people have emphasized the
27:40old age security hypothesis. There's a
27:43famous um you know demographer Caldwell
27:46John Kwell has written extensively about
27:48this. There's also some economists who
27:50followed up. I think there's probably
27:52something to that that um you know you
27:56needed kids as you know old age security
27:58and then as for example um pension
28:01systems were introduced or other ways of
28:04saving for retirement or for for old age
28:07and so forth there was less need of
28:10having you know many kids to secure your
28:12old age um you know well-being um but I
28:16don't think that can explain the whole
28:18um demographic transition as well so I I
28:20think personally think the most you know
28:23in reading all this literature and and
28:24working in this literature and so forth
28:27the most convincing story is the
28:30increasing returns to human capital
28:32because of some kind of technological
28:34progress leading people to want you know
28:38more educated kids and because of that
28:40fewer children
28:42um and so there's a famous paper by Gala
28:45and while in the AR 20 years ago that
28:48kind of develops this you So the
28:50hypothesis very clearly in a model and I
28:53think
28:55um yeah let me yeah I was going to show
28:58so there's a bunch of papers now looking
29:01at all sorts of aspects of the
29:03demographic transition that at the heart
29:06have this kind of mechanism increases in
29:09the returns to human capital leading
29:11people to want more schooling and
29:12because of that having fewer children
29:14sometimes it's a little bit hidden so
29:16it's not because the papers are maybe
29:18about something else as Well, but if you
29:21dig deep and ask the question why is it
29:23precisely that fragidility falls often
29:25this mechanism is at the heart. Um so I
29:28wanted to very briefly show in a toy
29:31model how that works and then mention a
29:33few papers that are sort of based on
29:36this hypothesis or on this mechanism uh
29:39which to me seems the most important uh
29:42and um and not just to me I mean I think
29:44there's a consensus now in the
29:46literature that that is the most
29:48important not only but quite important
29:51factor in explaining fertility decline
29:53over time uh and development so economic
29:57development increasing at the same time,
29:59right? It really connects the two. Um,
30:02so here's a very basic toy model just to
30:05show this in a static almost static yeah
30:08essentially static setting. Uh, if you
30:11want more elaborate versions of this,
30:13you know, Barrow Becker is a famous
30:15model that basically has this in a much
30:18more elaborate dynamic setting. uh but I
30:20think to make the basic point is enough
30:22to look at this very simple just
30:25parental decision problem where a parent
30:27cares about own consumption UFC cares
30:31about the number of children n uh and
30:34also the well-being of the children so y
30:36prime is supposed to be so y think about
30:38is output and y prime is output in the
30:41next period um so then the parent has to
30:45choose how much l um how much um to work
30:50Um how
30:53yeah
30:55and then yeah this these are the costs
30:57of children. So n is a number of
30:59children and children costs both uh five
31:02is just some fixed cost. This is how
31:04much children cost and e is education
31:06time. So the way I wrote it down here is
31:09how much time parents spend you know
31:12educating their own children. So this is
31:14not public education. This is parents
31:16spending time with kids educating them.
31:19and then output the next period. Um
31:21yeah, so I I misspoke when I said static
31:23model. It's a very simple, very stylized
31:26model, but it's not static. It is
31:27dynamic in a sense. This is the problem
31:29of one parent, but then time continues.
31:32Um human capital accumulates over time.
31:35Uh and there will be growth. There is a
31:37growth model in that sense kind of a
31:38very simple AK AK model I wanted to say
31:41but ah model actually. So h is human
31:43capital here and human capital um is
31:46increased through this education time
31:49and it's a very simple functional form
31:51here. So with that you can actually uh
31:54solve for closed form solutions if we
31:56also assume lock utility.
31:58Then the optimal education in this model
32:01is just this term here and it's
32:03specifically a function of the returns
32:07to education the theta right these the
32:09theta are the returns to education that
32:11enters but also how much you care both
32:13about the quality of your kids and the
32:15quantity of your kids that's the these
32:18two gamas that's the quantity you know
32:20how much weight you put on quantity and
32:22this is the weight you put on quality
32:24right the well-being of your children
32:27if If you then think the theta the
32:30returns to education have increased over
32:33time then with this symbol model you can
32:36explain the demographic transition.
32:38You'll have the human capital right the
32:40growth this the growth rate of human
32:42capital it will increase as the theta
32:45increases and then go back to the
32:48production function right so output so
32:50what people are eating is just direct
32:53function of the human capital with some
32:55TFP factor in front of it and so as
32:58human capital increases output increases
33:01in the exact same way um so just this
33:04increase on theta uh leads to more
33:06education uh uh and it leads to um more
33:11output at the same time it leads to more
33:14not more sorry less fertility. So yeah,
33:18I should have put that on the slide. So
33:20I guess Suzanne can show that the n the
33:22optimal n um is going to be a similar
33:25well you can also solve it in closed
33:27form and there you will see as data in
33:30theta increases fertility will fall. So
33:33education and fertility will go in
33:35opposite directions and there you go
33:37done with the demographic transition in
33:39in one slide.
33:42So the key of course is the schooling or
33:45the education in that story. So what
33:48does the data show us about that? So
33:50here again in cross-country data uh
33:53expected years of schooling so a measure
33:55of schooling completed schooling and the
33:58total fertility rate are also very
34:00strongly negatively related in
34:02cross-country data.
34:04So that at least fits.
34:07And here I wanted to just very briefly
34:09mention that you know the papers in
34:11which this basic mechanism is present
34:13there's just lots of you know macro
34:16development papers that are at the heart
34:19if you dig deep essentially you'll find
34:20the version of this mechanism um you
34:23know an older paper by Becca Murphy and
34:25Tamura um then another paper by well the
34:29original paper by Galoran while they
34:30have another paper has the same
34:32mechanism at the heart uh a paper on
34:35mortality by Zoraris
34:37in Greenwood and Sashadri's paper the
34:39focus is more on agriculture and
34:41urbanization but it's sort of a you know
34:43quite connected
34:46um even if you look at this paper by
34:48Manuel and Sadri they're trying to
34:50explain international fertility
34:52difference today uh or you know I don't
34:55know 10 years ago maybe um even there um
34:58the basic mechanism it's a very
35:00complicated model but the basic
35:02mechanism I would argue is sort of this
35:04you know this on. Um,
35:10so any questions so far?
35:14Okay, everyone is so quiet. Maybe it's
35:18maybe it's too simple. So, let's move on
35:20to something more complicated.
35:22So, the focus so far, everything I've
35:25shown you is on the tradeoff between or
35:28the decision between parents and
35:29children in one gender model. There, you
35:32know, there's no two parent. Have you
35:33seen two parents anywhere? No, all of
35:36these models are about just parents.
35:39Parents, what does it mean parents?
35:41Usually, I mean, there are at least
35:43biologically everyone has two parents.
35:46Um, great. So, let's think about the
35:50family a little bit more seriously and
35:52why would that matter and how can we
35:54think about that in models but also for
35:56what kind of questions does it matter?
35:59So, a family right typical family there
36:01would be a husband, a wife, there would
36:03be sons and daughters. So multiple
36:04family members um fathers and mothers
36:08may actually disagree about the
36:10investment into children. Um if so then
36:13bargaining in a family might become
36:15important and there's some recent papers
36:17exactly about this. Um that's about the
36:22preferences but men and women or mothers
36:25and fathers may also enter the
36:26production function differently. Um
36:30right so they might not have the exact
36:32same skills. um or they may interact
36:35with kids differently, right? Or just
36:37the fact that there are two may actually
36:39matter for child production, child human
36:41capital production. Um mothers and
36:45fathers may also enter differently in
36:46just the pure human capital transmission
36:48part. So if you think there's some
36:50learning by doing by just watching your
36:52parents, uh you know, watching one
36:54versus the other may make a difference.
36:57um in some context um you know families
37:01may have a son preference um I mean
37:04there's a large empirical literature on
37:06that that seems to be case at least in
37:08some seems to be the case at least in
37:09some settings
37:11um and another aspect of family sort of
37:14to what extent are things that are being
37:16consumed in the family public goods
37:18versus private goods um yeah so all of
37:22these things cannot be thought about you
37:26cannot think about them in a model with
37:27just you know one you know one parent
37:30families.
37:33So well here's a super simple model in
37:36or well super elaborate a little bit
37:38more elaborate but still relatively
37:40simple model um in which there are
37:44basically two parents husband and wife
37:46or female male and female
37:49um and I'm allowing for lots of
37:52different differences between them just
37:54to kind of show how to add that and
37:56still you can solve this in closed form
37:58solution that's the beauty here um so
38:02one is the utility function, how much
38:06they care about the well-being of a
38:08child may actually matter. So maybe
38:11women might be putting more weight on
38:13kids than men or vice versa. I'm not
38:15taking a stand here. I just want to show
38:17what would happen if you model it this
38:19way. You can also think about the
38:22bargaining power or the say of each
38:25parent differing. So that's sort of the
38:28weight the bargaining weight on the
38:29woman and that's the bargaining weight
38:31on the man.
38:33Then what about the um you know how much
38:37they enter into production into this is
38:39meant to be sort of market production if
38:41you want right how we're con producing
38:43final consumption goods um you know
38:46maybe maybe women and men are complete
38:48perfect substitutes in that world I mean
38:51here I'm allowing at least um to have
38:54them not be perfect substitute
38:57um then the education for sons and
39:00daughters that's the education time here
39:02again you know in principle it could be
39:03different you could be spending more
39:05education time with you know daughters
39:06and with than sons or vice versa again
39:10um this is the
39:13human capital transmission part right on
39:17the one hand so human capital of the
39:19kids is determined through a how much
39:21time you put into them education time
39:23schooling but also this is what I called
39:25before learning by doing this is sort of
39:28just that your own human capital if you
39:31uh enters into theirs. It's they they
39:33see what you're doing and somehow they
39:35learn from that. And in principle, it
39:37could be again that mothers and fathers
39:39differ, right? So if beta was 1/2, it
39:41was can be completely symmetric. If beta
39:44is larger than 1/2, the females matter
39:46more and vice versa. Um and
39:52yeah, that's well and then that's the
39:54production of the kids in the next
39:56period, right? So that's essentially the
39:57same as the C here above just for the
40:00next periods. Um and then let's just
40:03assume all kids marry. And since
40:05everyone all men are the same, all women
40:07are the same, there's no heterogeneity,
40:09matching is irrelevant here. Um so then
40:13again under these assumptions kind of
40:15simple model lock utility, you can solve
40:18again for the equilibrium growth rate.
40:21Um and uh here I'm showing you sort of
40:24what it is just to think through which
40:26parameters matter. Um and I'm telling
40:29you now what determines growth but in
40:31this model and again Suzanne can go
40:34through some more details uh in the TA
40:36section but in this model the higher is
40:37the growth rate also the lower um you
40:40know or in a little bit augmented model
40:42the lower would be the fertility rate
40:44right as you see I actually didn't put
40:46fertility here now I just put this is
40:48just about the human capital of the
40:50children but basically you can marry
40:52this or combine this with this simple
40:56model where the quantity of children
40:58also entered. So that would be even more
41:00complicated. So that's why I didn't do
41:02it. Uh but the same mechanism would
41:04apply here. And so then you see
41:07fertility would fall by you know through
41:09similar mechanisms.
41:12But basically what I'm trying to the
41:14point I'm trying to make here this
41:16equilibrium growth rate now
41:19uh is determined by a bunch of different
41:21things not just by the growth rate in
41:23education but also for example the
41:26relative bargaining power of you know
41:28men versus women matters that that was
41:30the lambda um and you know the weight
41:35people put on their own kid that
41:36increases the growth rate. So you see in
41:39some sense how much the family matters
41:41or details of the family would matter
41:43for for you know growth and development.
41:48So yeah this logic sort of the two
41:52gender logic is again you know happen or
41:56is is you know relevant is important for
41:58a bunch of you know recent papers. Uh I
42:02just wanted to again mention a few here.
42:05So um this Gallorian wild paper um well
42:08that's not quite so recent but um this
42:11it's called from bra to brain um the
42:14decline in the gender wage gap is there
42:16the key force for the fertility decline.
42:19So you know for to make that point you
42:21need a two gender model. Um there's an
42:24older paper by myself um where I'm
42:27showing how polygyny so the fact that
42:29men can marry multiple wives lowers the
42:32capital stock and increase fertility.
42:34And so again to even think about
42:36polinigy in in you know you need to
42:39gender. um a recent paper that I find
42:43very exciting that just came out in
42:44econometrical two years ago now by
42:47Chhurst Jones and Cleo they show how the
42:50improved allocation of talent across
42:52gender and also ethnic groups but in the
42:55end that doesn't play a huge role mostly
42:57gender accounts for a sizable chunk of
43:00the growth you know a measured growth in
43:03the United States over the last 50 or 70
43:05years now um so again you need two
43:09genders to to even think about that. Um,
43:13in another recent paper that I have with
43:15Matias Dupka, we're looking at transfers
43:19to women. So, development aid given to
43:22women. To what extent that may reduce
43:25growth and um, you know, for that we
43:28have a model with endogenous division of
43:31labor in a non-cooperative family
43:33context. Again, you need two two you
43:36know, two family members um, to think
43:38about that.
43:40And then last and I will actually spend
43:42you know a lot more time on that now is
43:45you know thinking about um the
43:48importance of you know female bargaining
43:50power for uh leading to higher growth
43:52rate and possibly for to the evolution
43:55of women's rights again connected very
43:57much connected to demographic transition
43:59because the force the key mechanism will
44:02be going through fertility decline. Um,
44:07okay. So, with that, let's take a little
44:09bit closer look again at gender and
44:11development. Now, um, unless there are
44:14questions,
44:16Michelle, there are a couple of
44:17questions and, um, a couple of
44:19clarifying questions that, um, came up.
44:22One is just a factual question about the
44:25decline in fertility and the decline in
44:27infant mortality. Do those things match
44:30in terms of timing? Um, I guess that
44:33would be the first one. And then some
44:36more speculative things about the role
44:37of of culture, of religion, um, and and
44:42of course the issues I think you're
44:44about to get into, but about gender
44:46norms and women's agency and so on. So
44:48those I suspect you'll be coming to, but
44:50maybe just the factual question about
44:52the decline in infant mortality.
44:54Yeah. So there is a big actually debate
44:57even about the facts is my
44:59understanding. So it in some countries
45:02it seems to line up well. Mortality fell
45:05first and then fertility declined after
45:08and based on that some historians have
45:10argued for example there seems to be
45:12close connection and then in other
45:14countries and other regions um fertility
45:16fell first so then it seems you know not
45:19so connected. Um I've seen a bunch of
45:22empirical papers on that people seem to
45:24be digging up all sorts of kind of data
45:26and and keep arguing about it. So I'm
45:29not 100% sure. I think for some regions
45:32there seems to be a good you know
45:33connection in terms of timing but not
45:35everywhere.
45:37Um
45:39yeah
45:40okay and then yeah the other questions
45:42were about culture and gender norms. So
45:44let's think about that. So yeah exactly
45:47I want to talk about gender and
45:48development now. Um
45:52so this is the um gender empowerment
45:55measure by the OECD. So it has a bunch
45:57of different indicators of um yeah of
46:01gender of constraints on women um but
46:03also outcomes a little bit of a mix of
46:06um you know cultural constraints
46:09differences in inheritance rights for
46:11example um but also differences in
46:14schooling which is more of an outcome of
46:16course but uh looking at that indicator
46:19and GDP per capita you see again a
46:21strong positive relationship right so
46:23it's the richer countries in which women
46:26have more rights. Um
46:29then I wanted to just show you just very
46:31briefly a few other specific indicators
46:35which I find a little bit more
46:37interesting in a sense I think to me
46:39they're more a direct measure of
46:41constraints so rather than outcomes
46:43right if I'm just looking at female
46:45labor force participation or female
46:47education there could be all sorts of
46:49economic forces driving people to make
46:50these choices so I think from an
46:52economic point of view it's good to have
46:54sort of direct indicators sort of to
46:57what extent are men and women treated
46:59differently by law, for example. How are
47:01the constraints different? And so access
47:03to land is one such measure. I mean,
47:08you know, people put a lot of work
47:09trying to, you know, put all sorts of
47:12different laws into an indicator. Now,
47:16I'm sure you can in argue with them how
47:18well it's done, but let's assume they
47:20did the best they can. Um then again we
47:23see here
47:25um sort of that access to land among the
47:28rich countries um you know basically
47:31they all have almost full access to land
47:34uh and but then you know it is downward
47:36sloping um here and but you know there
47:40also a bunch of poor countries here down
47:42here that still that supposedly have
47:44full access too. Um so it's yeah but no
47:48rich country you find you know here
47:51right so in that sense a pretty strong
47:54relationship
47:55and yeah and if if anyone is interested
47:58in in more of these indicators um in a
48:01survey that I wrote with Alexandra Vena
48:03and Matias Dupka 10 years ago uh we put
48:07a lot of these indicators in the survey
48:09it's a long appendix um and also gives
48:11all the sources in the annual um review
48:14of economics
48:15Um yeah so here's another measure called
48:19parental authority. So to what extent um
48:23male and female have more say you know
48:25husband and wife have more or mothers
48:27and fathers have more say uh towards
48:30their children. Um so there's a gender
48:33bias also for for a bunch of countries.
48:36Again if it's one then they have the
48:38same say and again basically no rich
48:41country do you see you know with very
48:44asymmetric parental authority indices um
48:49okay so there is and a bunch of other
48:52measures I just wanted to highlight
48:53these two as examples
48:56um but there are so basically a bunch of
48:58different papers um or you know or at
49:01least empirically it's very clear that
49:03is a positive relationship between
49:05women's rights and development. Now, you
49:08know, interesting question is why would
49:10this be the case? So, one logic could be
49:13just to think about more rights, for
49:14example, more property rights can lead
49:17to higher returns
49:19and that leads to more investment. And
49:22there's a bunch of empirical papers
49:23looking specifically at this channel. I
49:25don't want to go through all of them
49:26here right now. Um, but I think usually
49:30the effect is pretty small. So I think
49:33I'm you know if you give women more
49:35rights they will leave on the margin to
49:38higher output. I think there's no doubt
49:40that's what these papers document but I
49:42don't think that's the main cause for
49:44development. Right. Um
49:47another argument um that's made as sort
49:51of more political rights for women can
49:53lead to better government spending and
49:55that can lead to more efficient
49:56outcomes. So again I listed a bunch of
49:59papers here that go in that direction.
50:01Um, oh yeah, and if you're wondering
50:03about all these papers, if anyone is
50:05interested specifically in one, I know
50:07the reading list I gave you was short. I
50:09didn't include them all, but Suzanne
50:11actually just put together a complete
50:13reading list that sites, you know, that
50:15includes all of these papers on these
50:17slides. So, if you're interested in a
50:18specific one, um, I think she'll hand
50:21that out in the TA section or we can
50:23also email it around. Um, but yeah,
50:26we'll have a complete reading list if
50:27you want to look up any of these papers.
50:29um yeah
50:32okay so more political rights better
50:34government spending um and that can lead
50:37to some increases and output but again I
50:39don't think that's the main driver of
50:41development
50:42um actually skip this one
50:47so I think the more interesting
50:49relationship is the other way around
50:51that it's the development process
50:53leading to more women's rights
50:56and so then we can think about the
50:59family as a driver of political change,
51:01right? So women's rights is largely
51:03determined by politics, right? As
51:05politicians who decide on laws uh and so
51:09to what extent um you know are these
51:12laws changed and when are they changed
51:14and why are they changed? So let's think
51:15a little bit more about this political
51:17economy aspects now. um
51:22right so if if you take a step back if
51:25you think during the development process
51:28basically all of today's rich countries
51:30went through a similar series of reforms
51:33I would say right democracy spread
51:35everywhere
51:37introduction of public schooling
51:38happened pretty much everywhere child
51:41labor was banned everywhere welfare and
51:44social security systems were introduced
51:46in in all countries and women's rights
51:50were introduced sort of you know in
51:52sequence not all at once but as
51:54countries got richer they basically went
51:56through all of these reforms
51:59so the questions I want to raise
52:00basically in the last 20 minutes um why
52:04did these reforms happen when they did
52:07why did they fail to happen in some
52:09countries uh what do economists have to
52:12say about this and again with the
52:14connection to to the de demographic
52:16transition of course um Right. So here's
52:20some examples
52:22again some papers. So Galor and Mov have
52:25an interesting paper on the role of
52:27capitalists in the expansion of public
52:29education. So it was precisely because
52:34uh you know the capitalists people
52:37running factories they wanted more
52:38educated um workforce and so they were
52:42partly responsible for not not
52:45exclusively but that's the point of the
52:46paper for sort of the expansion of the
52:48public education system.
52:51Another super interesting paper by um
52:54Dupk and Silibati is the child labor
52:57bands. So the point they make in that
53:00paper is that over time and this is very
53:02much connected to the demographic
53:04transition over time as returns to
53:07education increased people wanted more
53:11educated children and then it became
53:14beneficial to ban child labor. Um this
53:18story is more subtle. I mean it's hard
53:19to summarize in a few sentences. Um
53:22there is a trade-off. Basically, if you
53:24think about it from a perspective of a
53:26parent, children
53:28um compete with you. If you're an
53:30unskilled worker, child labor competes
53:33with you in the labor market. So then
53:36it's maybe better not to have them work.
53:39But if you're a very poor family, well,
53:41you need the children to work otherwise
53:43you're starving. So there's this
53:45trade-off. on the one hand the
53:47competition on the other hand is sort of
53:49income for the children from the
53:52children for the family and so they show
53:54there is a or they have a model they
53:56show you know at some point when returns
53:59to education are large enough people all
54:01want fewer children more educated
54:04children then at some point it pays or
54:06the majority becomes in favor of child
54:08labor bans
54:10um and then the last example and the one
54:13I want to spend some more time on
54:15because This is one of oops one of my
54:17own papers is uh thinking about the
54:20expansion of women's rights. Um and so
54:23there we want to think about the
54:24incentive of men to vote for women's
54:27rights. But all of these are examples of
54:30you know that argue basically that
54:32technological change led to some changes
54:35in the family
54:37um that made you know reforms desirable
54:41which in turn increase output and
54:42growth. And by and large they're all
54:44related to fertility declines at the
54:47same time, right? Increases in education
54:49for the kids, smaller families. Uh and
54:53in that context, these reforms make
54:54sense.
54:58So yeah, this was just a picture in
55:01terms of the child labor. I don't want
55:03to um spend much time on that. Again,
55:05it's it's this very exciting paper by
55:07Tiliboti and Duckka, but the child labor
55:10rate is highest and this is data from
55:141990, so it's 30 years old, but 30 years
55:17ago still, um the child labor rate was
55:20the highest precisely in those countries
55:21where fertility was very high. Uh and
55:24then in countries where fertility, you
55:26know, was more much lower. So, you know,
55:28anywhere, you know, below three or so,
55:30here basically there is no child labor
55:32anymore. Um
55:36okay any more questions
55:43so then I want to now talk about the
55:46evolution of women's rights as sort of
55:49another example of you know reforms in
55:52the family or you know changes in the
55:54family that led to political reforms and
55:57that at the same time sort of caused or
56:00contributed even further to the
56:02demographic transition.
56:06Right? So the the question here is
56:09basically if look if you look at
56:11developed countries there were dramatic
56:13changes in uh women's rights and
56:16specifically married women's rights over
56:18the last 200 years. If you think about
56:20200 years ago, basically even in the
56:23United States, in Germany, in the UK,
56:25women essentially had no rights uh
56:28especially married women upon marriage,
56:30their legal identity essentially merged
56:32with the husband. They sort of ceased to
56:35exist as a legal entity. Um and then
56:39lots of changes uh a lot of them in the
56:41second half of the 19th century. Divorce
56:44law, for example, um came into place and
56:47made it possible at least to some extent
56:50for women to separate from abusive
56:52husband.
56:54Child custody law, very interesting, was
56:56reformed towards in favor of women. So
57:00nowadays, you know, most recent changes
57:02maybe go in the opposite direction,
57:04right? I think right now most child
57:05custody laws are quite biased towards
57:07women actually and it's quite hard for
57:10fathers to to gain access to a child.
57:13But if you go back in history, it was
57:14the opposite. Children belong to the
57:16fathers. And so a lot of women were
57:19actually staying even in horrible
57:21marriages, abusive relationship because
57:23they knew if they left the husband, they
57:25would lose the kids and they didn't want
57:26that. So then there were changes in
57:28these child custody laws making it
57:30easier for women to have access to their
57:32children upon separation. Uh and that
57:35happened basically in the second half of
57:36the 19th century. And then very
57:39importantly also married women's
57:40property law. So basically you know
57:43women used to not be able to really own
57:45anything separately. Uh so then again
57:48how do you want to you know separate or
57:50how do you want to make own decisions or
57:52you know do anything if if you can't
57:54have own anything. Um
57:58okay so that was changed again. Um so
58:01but initially all of these increases in
58:04women's rights um you know basically
58:08were done by by men in the sense right
58:11all the parliaments all the people in
58:13power either if it was fully democratic
58:16I mean basically the voting was done by
58:19the men the politicians were all male um
58:23and yet it's these male men who decided
58:26to share you know to give these rights
58:28to women so it seems an obvious question
58:31why did men do that? Um
58:34so the approach here we took in this
58:36paper is a formal model of women's
58:38rights and the focus again is on the
58:40family. Um right so the expansion of
58:44female rights started long before the
58:46widespread increase in female labor
58:48force participation. So I don't think
58:50it's it's very logical that it's
58:53connected to women working.
58:56Second point, it's large changes in the
58:58rates of married women specifically. So
59:00it's nothing to do with the single
59:01women. There was not so much changes. So
59:04that's I think again why a focus on the
59:06family makes sense. Um and these
59:09expansion of rights really coincided the
59:12time and coincided with these large
59:14fertility declines and the rise in
59:17education. Um and just to emphasize that
59:20again the political rights were granted
59:22to women later. So in the in the United
59:25States was um what was it 1919
59:28right or 1918 when the right to vote was
59:31implemented Germany around the same time
59:33most European countries even later more
59:35like right after second world war.
59:39So voting rights happen much later in
59:41most places.
59:47So here's the idea. Women's rights
59:50determine bargaining and marriage. If
59:52you cannot own anything, if you're gonna
59:55lose your kids upon divorce, if you
59:57can't even initiate a divorce, basically
1:00:00as a woman, right, you have essentially
1:00:01no bargaining power. The husband can
1:00:03decide everything. If you, you know, if
1:00:06you could divorce, if you can take your
1:00:07kids with you, if you have have actually
1:00:09some property, then you know that shifts
1:00:12the bargaining.
1:00:15So then if you think about it from a
1:00:17man's perspective, there's a clear
1:00:19tradeoff between your own wife and other
1:00:22man's wives, your own wife, well, why
1:00:25should she have any bargaining power
1:00:27ever? She's just going to take something
1:00:29away from you, right? Or if I mean, if
1:00:31you like her, you can give her stuff,
1:00:33but for that she doesn't need bargaining
1:00:34power. Bargaining power is only relevant
1:00:36if you disagree.
1:00:39But if you like your daughters or also
1:00:43other you know mothers other you know
1:00:46people in the economy it can make sense
1:00:49from men's perspective that he wants
1:00:52other women in the economy to have more
1:00:54rights not his own wife but for example
1:00:57his daughters right if you don't want
1:00:59the daughters to be exploited by their f
1:01:02by your future son-in-laws it's maybe a
1:01:04good thing that the daughters have more
1:01:06rights um or maybe mothers of your
1:01:10children's future spouses. If you think
1:01:12that mothers are important for the
1:01:14education of kids, then you want mothers
1:01:17to have more say so they invest in their
1:01:19kids. So you have a better sort of
1:01:22spouses to pick from for your own kids,
1:01:24more educated, better um so then there's
1:01:28a trade-off and that tradeoff that's the
1:01:31point of this paper depends on the
1:01:34returns to education.
1:01:36how much you're losing visa v your own
1:01:38wife that's kind of a static tradeoff
1:01:42she has more bargaining power you're
1:01:43always losing some fraction let's say of
1:01:45your consumption or of your income um
1:01:48but how much you're losing in terms of
1:01:50thinking about your daughters or about
1:01:52mothers investing in their kids who
1:01:55might be spouses for your kids that's
1:01:57kind of a dynamic decision and again the
1:02:00higher the returns to education
1:02:03the more important is that motive So as
1:02:06the returns to education increase over
1:02:08time
1:02:10that's exactly that same mechanism that
1:02:12toy model I showed you at the very
1:02:14beginning right the theta was increasing
1:02:16leading people to want educ more
1:02:18educated but fewer children that exact
1:02:20same mechanism is at the heart of this
1:02:22paper as well so returns to education
1:02:25increase that changes you know the
1:02:28desire to education that flips the
1:02:31tradeoff and that makes men vote for
1:02:33more women's rights.
1:02:37So yeah, so in some sense that's the
1:02:40paper in a nutshell. Let me show you a
1:02:42little bit more about the model and just
1:02:44a computed example. Um, but I don't want
1:02:48to go into too much detail here, but
1:02:50it's an overlapping generations model of
1:02:52men and women. Um, all men and all women
1:02:56are identical here. So there's no
1:02:58interesting heterogeneity.
1:03:00Men and women are always joined in
1:03:02marriage. There's no marriage decision,
1:03:04right? The kids just marry other
1:03:06people's kids since everyone is the
1:03:07same. There's no question about
1:03:08assortative mating. It's quite simple in
1:03:10that sense. Utility as an toy model is a
1:03:14function of consumption of fertility and
1:03:17children's utility. People are
1:03:19altruistic towards their kids. Um so in
1:03:22that sense it builds upon this old
1:03:24barbeca model. So anyone who wants to
1:03:26think about fertility models and think
1:03:29about demographic transition more
1:03:30seriously needs to read the barbeca
1:03:32paper. I mean that is the classic in
1:03:34this literature. Um
1:03:38and then it's also an endogenous growth
1:03:40model in the sense that is human capital
1:03:42accumulation. Um decision making in
1:03:46marriage. Here we're analyzing two types
1:03:48of regimes. One is the pure patriarchal
1:03:51society. Men just make choices. That's
1:03:53it. The women have no say and a second
1:03:56regime equal power sharing. They make
1:03:58decisions jointly and then men only men
1:04:02can vote on the regime.
1:04:05Key assumption here is that mothers care
1:04:07more about the children's welfare than
1:04:09fathers do. So again in this toy model I
1:04:11had this um gamma and I allowed it to
1:04:15differ by gender right how much weight
1:04:17is put on children. So here the
1:04:19assumption is that women care more put a
1:04:21higher weight on the well-being of the
1:04:24children than than fathers. It is
1:04:26important for the mechanism here.
1:04:30So yeah let me maybe skip these
1:04:32equations. Um
1:04:36yeah let me see I don't have that much
1:04:38time so let me I mean it's very similar
1:04:40to the toy model except now we really
1:04:42have these two genders. they make these
1:04:44choices uh together and then it's the on
1:04:47top of it is this voting for what regime
1:04:49do you want to live in right so that's
1:04:52here this is the two regimes the
1:04:54patriarchal regime is just men making
1:04:56decisions they just solve their own
1:04:59value function the women just do
1:05:01whatever the men tell them to and the
1:05:03empowerment regime they just maximize
1:05:05the weighted sum or not even weight just
1:05:08the sum of the two uh value functions
1:05:12and the man just vote on the regime and
1:05:14you vote once and for all that will
1:05:16affect all the current and future
1:05:17marriages.
1:05:20And then the results here,
1:05:23if the returns to education are low, so
1:05:25it's the same theta basically, then
1:05:28parents, if they're really low, parents
1:05:29don't educate at all, the decision
1:05:31problem is essentially static.
1:05:33The political regime only affects the
1:05:36consumption sharing between husbands and
1:05:37wives. And so because of that, the man's
1:05:40incentive for sharing power is low.
1:05:42basically they would be losing a lot.
1:05:44The women started, you know, eating too
1:05:45much. They're not gaining anything
1:05:48because there's no point in educating at
1:05:50all. If the returns to education are
1:05:53high enough, then the dynasty starts
1:05:55accumulating human capital and now the
1:05:58political regime affects the speed of
1:06:00accumulation.
1:06:02And if it's sufficiently high that
1:06:04return, then the trader flips and the
1:06:07men prefer to share power. Um
1:06:11yeah, here's some more details. Um but
1:06:14maybe in the interest of time, let me
1:06:15actually skip this. Some more detail
1:06:17exactly why it happens. Um if there are
1:06:20questions about that, maybe also Suzanne
1:06:21can go some more details or later in the
1:06:23Q&A, we can talk about that. Um just
1:06:27very briefly, I wanted to show you a
1:06:28computed example of this transition to
1:06:31power sharing. So I'm just starting the
1:06:33economy in a no education regime and
1:06:36then I'm gradually increasing this
1:06:38parameter theta. Um and I'll show you
1:06:41what happens here. So this is I'm
1:06:44plotting here fertility and education.
1:06:46Um you could also plot human capital and
1:06:49GDP it will move in the same direction.
1:06:51So fertility falls, education will
1:06:54increase as you'll see and human capital
1:06:56and GDP will actually move just in lock
1:06:58step with the lower panel here with the
1:07:00education. So just keep that in mind.
1:07:03Right? So first fertility is high. Five
1:07:05kids um education is zero. No one is
1:07:08educating their kids. And I'm first
1:07:10showing you what would happen under
1:07:12permanent patriarchy. Right? So men will
1:07:14make choices forever. Yeah. Fertility
1:07:17falls. Education goes up. So there would
1:07:20be a demographic transition already just
1:07:22because the theta increases.
1:07:25But what would happen now if they
1:07:26started sharing power? That's the red
1:07:29line. Now
1:07:31well first they choose
1:07:34so this is optimal empowerment the
1:07:36optimal empowerment initially for low
1:07:38values of theta
1:07:40um you know is no empowerment so
1:07:42initially there's still it's the two
1:07:44lines are on top of each other the
1:07:46optimal thing is per is patriarchy but
1:07:49then as the theta increases and I'm not
1:07:52showing you here the theta but the
1:07:53assumption was theta keeps gradually
1:07:55increasing over time it diverges
1:07:59So meaning now it's optimal to give
1:08:01women more rights to switch to that
1:08:03other decision uh regime and as you do
1:08:07fertility declines further it was
1:08:10already declining but now it declines
1:08:12more steeply and education was already
1:08:14increasing but now increases more.
1:08:19Okay so that is the voting. So here in
1:08:22this example now precisely in period six
1:08:25is when you vote for optimal for for
1:08:27sharing power with women for more
1:08:29women's rights and at that point in time
1:08:31fertility declines further and education
1:08:33increases further
1:08:36right
1:08:37so in terms of timing implications
1:08:40fertility decline and the rise in
1:08:43education starts just before the
1:08:44expansion of women's rights but once
1:08:47they're extended the fertility decline
1:08:49and the expansion of education
1:08:50accelerate.
1:08:52Um so in principle something one could
1:08:54possibly test in the data. I don't have
1:08:56um formal tests here. It's not a very
1:08:58deep empirical analysis but I wanted to
1:09:02show you some pictures very briefly. Um
1:09:06this is just a timeline of female
1:09:07empowerment just to argue that by and
1:09:10large it was the second half of the 19th
1:09:12century where most of this was happening
1:09:14in the United States. So then let's look
1:09:17at education and fertility in the United
1:09:19States around that time. So the total
1:09:22fertility rate, that's the solid line
1:09:24here. Yeah, it was declining prior to
1:09:27the expansion of women's rights and then
1:09:29it kept declining afterwards. So it's
1:09:31right in the middle, right? If we're
1:09:33thinking about women's rights were
1:09:34expanded sort of, you know, between 1860
1:09:38and 1900, that's right in the middle of
1:09:41this fertility decline.
1:09:43And this is primary education rates and
1:09:46this is secondary education rates and
1:09:48especially the primary education rates
1:09:50they were rising big time just around
1:09:53that time as well and then they
1:09:54increased to pretty much 100 already by
1:09:571900. Um
1:10:00okay then we could look at this in
1:10:03England too but let me skip that in the
1:10:05interest of time but it's very similar.
1:10:08Um yeah so brief summary the extension
1:10:11of women's rights I think is the prime
1:10:14example of voluntary power sharing um
1:10:17voluntary as in through the political
1:10:19process. Um the theory explains why the
1:10:22rights were extended exactly you know
1:10:24when the importance of education changed
1:10:27really the role of the family and that
1:10:29mechanism again generates a demographic
1:10:31transition along the way um and an
1:10:34acceleration of the demographic
1:10:36transition. There is really a two-way
1:10:38interaction in this model between
1:10:40development and female empowerment. It's
1:10:42development first, right? Countries
1:10:44getting richer causing the female
1:10:46empowerment, but then the female
1:10:47empowerment feeds back into development
1:10:49as an output increases even more. Um, so
1:10:53one can think about implications for
1:10:54developing countries today. It's just a
1:10:56question. I don't have good answers.
1:11:01I let me just take one minute to um
1:11:06start um basically conclude this talk by
1:11:09with some thoughts on demography and
1:11:12development going forward. Right? So the
1:11:14demographic transition it has happened
1:11:16it's done tons of papers have been
1:11:18written about it. So you know maybe
1:11:21maybe we're done we can move on to new
1:11:24things possibly. So here are some
1:11:26related questions I think that have not
1:11:28been answered and that that are really
1:11:30important going forward for new
1:11:32research. Right? We can ask the question
1:11:35is lower population growth always better
1:11:37for development? Is it always better to
1:11:39have fewer people? And and is it always
1:11:41better for well-being for for utility
1:11:44for you know for yeah for well-being? Is
1:11:47there an optimal population size? You
1:11:49know, or what is the optimal population
1:11:51size? Or should we be shrinking in the
1:11:54limit? Um or or maybe not. Is shrinking
1:11:56a really bad thing? Um also, if people
1:12:00stop having kids or have fewer and fewer
1:12:02kids in most rich countries at least,
1:12:04we're below replacement rate right now.
1:12:06So, what you know, does that tell us
1:12:08about pension systems? Um how are we
1:12:10going to finance them in the future? Um
1:12:13or is there possibly an externality? you
1:12:15know, are we could it ever be we're
1:12:17having too few kids? Um, right. All the
1:12:20focus in the past has been on sort of
1:12:22the high fertility rates and maybe too
1:12:24high and you know, demographic
1:12:26transition being a good thing and that
1:12:28makes countries rich. Now going forward,
1:12:30you know, with super low fertility rates
1:12:32in in in at least parts of the world,
1:12:34you know, Germany 1.4,
1:12:36um, Korea 1.1 or actually one now. I
1:12:40mean, that's far below replacement. you
1:12:42need two kids to replace, you know, two
1:12:44parents. Um, right? Could this ever be a
1:12:47problem? So, I think it's worth thinking
1:12:49about this in that context. I want to
1:12:52briefly say that it's not so easy to
1:12:55even evaluate policies in this context.
1:12:57I just want to stress paral efficiency
1:13:00that's the concept economists usually
1:13:03use
1:13:05that's not defined in models where
1:13:07fertility is endogenous. par radio
1:13:09efficiency usually comparing alloca you
1:13:12know pe person by person across
1:13:13allocations well what if half the people
1:13:15are no longer there how you going to
1:13:17compare it uh it's just wanted be people
1:13:21to be aware of that in previous work we
1:13:24have um presented alternative concepts
1:13:26that could be used but I think so far
1:13:29not many people are using them so
1:13:32I think they're sort of useful but um
1:13:35but yeah I think not too many people are
1:13:38thinking about sort of too few kids or
1:13:41could there be really externalities in
1:13:43this context. Um
1:13:46so let me just take one more minute to
1:13:48highlight two reasons for why there
1:13:50might actually be too low fertility
1:13:53right now. I think knowledge
1:13:55externalities might be quite important
1:13:57here. So Chad Jones has a recent paper
1:14:00for whoever is interested. He makes the
1:14:02point if the discovery of ideas as a
1:14:05function of the number of people then
1:14:08parents will not fully internalize that
1:14:10the offspring create non-rival ideas for
1:14:13others and that could lead so he is
1:14:15saying in the limit the planet's going
1:14:17to shrink to zero and that's a bad thing
1:14:18basically that's what his paper is
1:14:20about. Uh we're going to be you know not
1:14:23growing anymore because we don't have
1:14:25enough people to to come up with new
1:14:27ideas. Um, and then in some other recent
1:14:30in another recent paper that I have with
1:14:32some co-authors. We're just kind of
1:14:34trying to finalize a draft right now,
1:14:37we're looking at South Korea
1:14:38specifically, but I think the mechanism
1:14:40might be relevant for other Asian
1:14:42countries as well. If comparison motives
1:14:45are important, as in you're comparing
1:14:47yourself with others, you want your kids
1:14:49to be super educated, to learn the
1:14:51violin, to know everything because
1:14:54you're worried about them falling behind
1:14:56other kids.
1:14:57If that comparison is super important,
1:14:59then parents may actually overinvest in
1:15:01the education of their children
1:15:04and then because of that underinvest in
1:15:06the number of the cats. So that could be
1:15:08driving the super low fertility rates in
1:15:11some Asian countries in Korea
1:15:12specifically and you know maybe um be
1:15:15also possibly a bad thing and bad for
1:15:18for economic um yeah well-being.
1:15:22Okay. And then I'm done. So let me just
1:15:25conclude by saying well I think families
1:15:27matter for the demogra well of course
1:15:29for the demographic transition they
1:15:30matter but they matter for the
1:15:32development process more generally.
1:15:34They're very important kind of part of
1:15:36the you know piece or piece of the
1:15:38puzzle. So then I'm done and happy to
1:15:41answer more questions or discuss the
1:15:43topics.
1:15:45Fantastic. Thank you so much Michelle.
1:15:48Um I'm going to open things up first to
1:15:50the to those sitting in the front row.
1:15:52you should probably just unmute
1:15:54yourselves and ask the questions
1:15:56directly and if we need to if I need to
1:15:58regulate traffic I can do that. There
1:16:00are also a few questions sitting in the
1:16:03Q&A that I'd like to pose if we have a
1:16:05bit of time. So who in the front row
1:16:08wants to wants to go ahead I see you
1:16:10have a hand.
1:16:11Okay. Uh well thank you Michelle for the
1:16:14very interesting lecture. Uh my my first
1:16:16question is with respect to the possible
1:16:18bicausal relationship between fertility
1:16:21rate and economic growth and
1:16:23development. Uh was it considered in the
1:16:27because I I I believe declining
1:16:29fertility could economic development
1:16:32could drive declining fertility and it
1:16:35could also be the other way around
1:16:37possibly. So how do we uh solve that in
1:16:39in a possible model? And um I read an
1:16:43article at a time looking at data from
1:16:45the United States of America uh talking
1:16:48about the effect of the evolution of
1:16:50digital technology on um uh resulting to
1:16:55rising leisure time and then less time
1:16:59for procreation
1:17:01as any think I mean people spend more
1:17:04time now on Netflix, social media and
1:17:06all of that than they spend on you know
1:17:09whatever and all of that. How do we
1:17:12factor this into uh all of this model?
1:17:15Thank you.
1:17:16Okay. Well, thanks for those questions.
1:17:17So, I guess the first one was about the
1:17:19direction of causality between fertility
1:17:21and development. Absolutely.
1:17:23Um I think so. I guess the the direction
1:17:27of causality I was trying to present
1:17:29here in a model is in some sense there's
1:17:30something else as in returns to
1:17:32education that sort of causes both if
1:17:35you want simultaneously.
1:17:37Um if you're just thinking about
1:17:39fertility sort of exogenously for some
1:17:42other reason cultural change or
1:17:44something falling and that causing
1:17:47development um I mean yeah it's hard to
1:17:52measure that um I know there were some
1:17:56attempts to you know through largecale
1:17:59birth control programs um in in India I
1:18:02believe um well obviously um the one
1:18:05child policy In China um there are a
1:18:09bunch of you know experiments if you
1:18:12want where basically people were forced
1:18:14to have fewer kids and that also that
1:18:17did lead to increases well depends on
1:18:20the context for sure in China I guess it
1:18:22led to increases uh in GDP per capita.
1:18:26It's not so surprising. I mean, if you
1:18:28have a given number of, you know, a
1:18:29capital stock and then you have you
1:18:31force people to have fewer kids than
1:18:33they want, sort of output per capita
1:18:35will increase as well. Um, but on a, you
1:18:39know, these were always kind of
1:18:40policydriven. And now there's questions
1:18:43again in countries like China, how do
1:18:44you fund the pension system then? Um but
1:18:48if you if you look historically in most
1:18:51countries that did not have such
1:18:53policies
1:18:54then I think it's pretty clear it wasn't
1:18:57that I mean why would fertility
1:18:58exogenously somehow change and that
1:19:01leading to development. Um I think there
1:19:04it's pretty clear must have been
1:19:05something else that caused fertility to
1:19:07decline. I I hope that answered your
1:19:10question. And then the second one was on
1:19:14um sort of different um leisure
1:19:17opportunities I guess um making people
1:19:19want fewer kids because they have
1:19:21alternative uses of time. Uh I think if
1:19:24you think about rich countries today
1:19:26that I mean is probably a reason why
1:19:29fertility
1:19:30is super low um in some countries. Is it
1:19:33the main reason or how do we measure
1:19:36exactly how much it contributes? I think
1:19:39it's an empirical question and I think
1:19:40it's out there. It's up in the air. I
1:19:42don't think it's been answered yet.
1:19:44Other questions from the front row.
1:19:47I would have another question. Um also
1:19:49thank you very much for your talk. Um I
1:19:52had two quick questions on the model you
1:19:54presented also on the voting which were
1:19:57um on like on this parameter about how
1:19:59much you care about your children. I
1:20:00think was gamma. I was just wondering um
1:20:04yeah how how do you estimate such a
1:20:07gamma or which numbers would you use and
1:20:10also like if you didn't want to assume
1:20:12that um women care more about their
1:20:15children than men. I I'm not sure we I
1:20:18think we didn't spend so much time on
1:20:19that slide, but wouldn't you achieve the
1:20:21same by just saying, for example, men
1:20:23just care more about the number of
1:20:24children than women do in the sense that
1:20:26maybe traditionally also when those
1:20:28voting rights were introduced, women
1:20:31just had to spend more time on each
1:20:33child dur like just just by yeah by by
1:20:37how society worked in a way.
1:20:39Yeah. So good question. So thank you. Um
1:20:42so in that particular model if you if
1:20:44you're asking about what's the size how
1:20:45to calibrate it how large should it be
1:20:48actually in that model we didn't
1:20:49calibrate anything it was about um
1:20:51showing a mechanism so it didn't matter
1:20:54so much what the number is um so we just
1:20:58needed a difference now you can ask sort
1:21:00of what's the micro foundation of that
1:21:03difference right why is it that men care
1:21:05maybe differently about the number of
1:21:08kids versus the quality and you're
1:21:10absolutely trade. All we need for that
1:21:12mechanism is that the quantity quality
1:21:14tradeoff is different for men and women.
1:21:17Right? So just assuming men care more
1:21:19about quantity will do the same thing.
1:21:22But leaving that aside, we can take
1:21:24another step back and ask you know sort
1:21:26of is there something deeper that causes
1:21:29that? And in an even more recent paper,
1:21:32Matias and I actually investigated that
1:21:34in a model where men and women have the
1:21:37exact same preferences
1:21:39and still they make different choices.
1:21:42And in that model, it's precisely um
1:21:45because of the non-corporative
1:21:46bargaining together with a gender wage
1:21:49gap that leads women I mean it's
1:21:52complicated. I cannot tell you the whole
1:21:54model in a nutshell, but we can get the
1:21:56exact same behavior that women choose to
1:21:59invest more in children because of an
1:22:02endogenous division of labor in the
1:22:04household where women are suddenly more
1:22:06responsible for the kids and men more
1:22:08responsible for earning a living purely
1:22:11driven by a gender wage gap. exogenous
1:22:13gender wage gap that leads women then on
1:22:16the margin to care essentially more
1:22:18about the kids um without assuming
1:22:21exogenous preference differences. So I
1:22:24think that could be sort of the deeper
1:22:25reason.
1:22:26Thank you a lot
1:22:27Michelle. I have a question from
1:22:29Elizabeth in the Q&A who says could men
1:22:32voting for women's rights also be
1:22:34explained by higher family income if
1:22:36both husband and wife have more equal
1:22:38access to the labor market rather than
1:22:40just the husband? Yeah. So that's a good
1:22:42point and good good question. That's
1:22:44precisely the point of that paper by
1:22:45Gettison and look it was somewhere on
1:22:47the slides. It's an empirical paper but
1:22:50they make that point that is precisely
1:22:52because of the agency cost. Um giving
1:22:55more rights to women would increase
1:22:58their incentive basically to to work
1:23:00well uh and that could lead to higher
1:23:02family income. uh and I you know they
1:23:04have data they claim you know this is
1:23:06convincing and I don't want to argue
1:23:08there's nothing to that story but by and
1:23:10large the timing is a little bit off um
1:23:14these rights or depends on which rights
1:23:15we're talking about but the the rights I
1:23:18was focusing on by and large were
1:23:19extended in the United States in the
1:23:21second half of the 19th century that was
1:23:24a time in which women specifically
1:23:26married women were not working in large
1:23:28numbers yet that happened much later so
1:23:31I think just the timing is a little bit
1:23:32off there.
1:23:33Other questions from the front row?
1:23:35Otherwise, I can keep throwing them at
1:23:37you from the Q&A. There's there's an
1:23:39interesting bunch of questions and um
1:23:42and
1:23:45um so Daniela Vidart says, "Is there a
1:23:48connection between human capital returns
1:23:51increasing and female labor force
1:23:53participation?
1:23:55If the returns to human capital increase
1:23:57for girls, it must be because there are
1:24:00or will be increasing labor
1:24:02opportunities for women and particularly
1:24:04educated women down the line. It
1:24:06therefore makes sense that the increase
1:24:08in female labor occurs later when these
1:24:10girls are grown. But the human capital
1:24:13returns increasing should come in hand
1:24:15also with female labor opportunities. So
1:24:18I think this is a question about timing
1:24:19in part um and and a question about what
1:24:22the relationship is between returns to
1:24:25education in the present and labor force
1:24:28opportunities in in the future.
1:24:30Interesting questions in there. Um
1:24:33empirically. So I guess one question is
1:24:35sort of empirically right what are the
1:24:37returns to education and how has that
1:24:40even changed over time is really hard to
1:24:44measure. I mean I'm convinced it has
1:24:46increased but I also want to take a very
1:24:48broad measure of inc um human capital
1:24:51returns. So for example um if um public
1:24:56you know from I want to measure from the
1:24:58perspective of a parents if I'm having
1:25:01you know as a as a country implementing
1:25:03public education that makes education
1:25:06much cheaper from a parents perspective
1:25:08you know that is another way of also
1:25:10from the parents there's returns to
1:25:11higher returns to education um but
1:25:14that's not something I can just measure
1:25:16by looking at um wages um relative to
1:25:20years of schooling So it's I want to
1:25:23think about a much broader measure of
1:25:25returns to education. Sort of how much
1:25:27does it pay for uh parents to invest in
1:25:31their kids. Um assuming there would be
1:25:33all sorts of government expenditures as
1:25:35well and so forth. Um so then it's
1:25:38really hard to have good long-term time
1:25:41series on this. So it's kind of a
1:25:44measurement problem. So then I guess the
1:25:47second part of the question was on um
1:25:49how has that then increased the female
1:25:52labor force participation. Um for that I
1:25:55guess it's the gender wage gap again is
1:25:58is key right? So in some sense returns
1:26:00to human capital accumulation is going
1:26:02to be much lower for women than for men
1:26:04if they're earning much lower wages in
1:26:06the labor market. Uh the gender wage gap
1:26:09has also declined um you know
1:26:11substantially over the centuries. In
1:26:13fact, Claudia Gold has a very long time
1:26:15series of the gender wage gap going back
1:26:17basically 150 years and you see it's
1:26:20been shrinking a lot um over time. So
1:26:23that of course I'm sure has played some
1:26:25role in the increase in female labor
1:26:27force participation as well. Uh and then
1:26:30how is that feeding back into fertility?
1:26:32I think that's especially important in
1:26:34the second half of the 20th century um
1:26:37or maybe starting or yeah right after
1:26:39the second world war
1:26:41um sort of women working more and more
1:26:43and more and fertility still declining
1:26:45somewhat um for quite a few decades.
1:26:49Yeah, I hope that answered the question.
1:26:53Any others?
1:26:54We have a fairly quiet front row, it
1:26:56seems. We're also coming close to the
1:26:58end and running out of time. So maybe
1:27:01I'll maybe I'll take advantage of a
1:27:02question posed by Francis um in the Q&A
1:27:06which is a broad question about um
1:27:09changing forms of marriage um and maybe
1:27:12I'll pose it as a question about how
1:27:14this literature is dealing with the on
1:27:18the one hand within primarily within um
1:27:21rich countries the expansion of marriage
1:27:24into same-sex marriage into civil
1:27:26partnerships of different kinds um
1:27:29changing forms of the household, but of
1:27:31course also thinking of the broad
1:27:33diversity of family and household
1:27:35arrangements that you've written about
1:27:37in terms of polygyny and elsewhere. So
1:27:39many of the models seem to seem to
1:27:41suppose um a husband a male husband and
1:27:45female wife bargaining um in a very
1:27:47simple family structure. I wonder if you
1:27:50can talk a little bit about how the
1:27:51literature is thinking about the broader
1:27:52diversity of family types.
1:27:54Yeah, that's a very good question. Um so
1:27:57first I want to say that much of the
1:28:00literature has no families whatsoever. I
1:28:03would say 90% of the demographic
1:28:06transition economics literature as just
1:28:08a parent. So that I think is already a
1:28:11step forward to think about two gender
1:28:13families a husband and a wife. That's
1:28:15already you know big step forward in my
1:28:17view and that allows us to answer a much
1:28:20richer set of questions. But then now
1:28:22the question um you know the question
1:28:24posed was about well what about other
1:28:26types of families there obviously the
1:28:29reality is even more diverse. Um and I
1:28:32totally agree and um the demographic
1:28:35transition literature I think has
1:28:36nothing to say about that but there are
1:28:38um a bunch of economists now working
1:28:40also on richer forms of families both
1:28:42empirically um but also in models. So um
1:28:47well some I think you mentioned
1:28:49polygyny. So I have a paper on precisely
1:28:51on that sort of looking at polygyny
1:28:53versus monogamy and how that matters for
1:28:55development. So that's one example but
1:28:58obviously I'm not the only one. There's
1:28:59a paper called the mystery of monogamy
1:29:02by um golden mo I believe. Um so that's
1:29:06specifically on polygyny versus
1:29:07monogamy. Then I think there was a
1:29:10question on samesex marriage. Um, so
1:29:14there I haven't seen all that much to be
1:29:16honest. Um, but I
1:29:19economically I mean this is about two
1:29:21people matching whether it's the same
1:29:23gender or opposite gender. Maybe from an
1:29:26economic point of view at least for some
1:29:27question doesn't make a big difference.
1:29:29Um, but empirically I think there are a
1:29:32bunch of papers on that as well.
1:29:34Theoretically I don't think I've seen
1:29:35much. And then the last thing maybe is
1:29:38um single parent. Well, one one question
1:29:42you could ask about really marriage
1:29:44versus cohabitation. Clearly,
1:29:46cohabitation has been on the rise in
1:29:49many European countries. There's
1:29:50actually now more cohabiting couples or
1:29:52as many cohabiting couples as married
1:29:54couples. But it to me that's kind of
1:29:56more a legal distinction. Economically,
1:29:59whether it's cohabitation or marriage, I
1:30:01mean, for all practical purpose, two
1:30:03parents living in the same household
1:30:04together. I don't care what it's called.
1:30:07So most or many economists lump
1:30:10cohabitation and marriage in the same
1:30:12bucket. Of course then there's divorce,
1:30:14there's single parenthood, but again
1:30:17there there are papers now looking at
1:30:19precisely these things. Jeremy Greenwood
1:30:21is a good example. Ny Gunnar has worked
1:30:24on that. Raquel Fernandez. So there's
1:30:26there's a lot of macro not a lot but
1:30:28there are some macroeconomists using
1:30:30richer versions of such models to think
1:30:33about precisely rise in divorce the rise
1:30:35in single parenthood premarital sex um
1:30:39and yeah other forms of of of families
1:30:42as well. Exactly. So
1:30:44sorry on a on a lighter note please. I
1:30:46just want to find out um uh your
1:30:49literature on um uh female empowerment
1:30:53towards the last graph that was a that
1:30:55was a graph where you graduated the
1:30:57increasing right and all of that. Did
1:30:59you consider any possible case of
1:31:02absolute and complete right? Maybe the
1:31:06world would be a better place if we give
1:31:07the women all the powers. [laughter]
1:31:13[gasps]
1:31:14Um
1:31:15I don't know that I guess that's a
1:31:16normative question. [laughter]
1:31:20Okay. Thank you very much.
1:31:21I think we should probably on that note,
1:31:23which is a a very nice note to stop on,
1:31:26but I just want to um thank Michelle so
1:31:29much for taking the time to to speak to
1:31:32us. I wonder if those of you in the
1:31:34front row would do what we've been doing
1:31:35recently and unmute and give her a round
1:31:38of actual applause. So,
1:31:43thank you very much. We will we will
1:31:46meet again tomorrow in the same time
1:31:48slot and uh Melanie Morton will be will
1:31:51be lecturing. So, look forward to seeing
1:31:54many of you there. And uh thanks again
1:31:56Michelle. really wonderful, clear and a
1:32:00great lecture that introduces some
1:32:01really important material for us. So,
1:32:04thanks so much.
1:32:05Thank you. Thanks for the opportunity
1:32:07and thanks everyone for asking
1:32:08questions. So, it was fun.
1:32:10Thanks so much.
1:32:11Bye.
1:32:12Bye for now.

About this transcript

This page contains the full transcript of Supplemental lecture: Demographic transition and development by CEPR & VideoVox Economics, generated from the public captions YouTube serves with the video. The transcript has 15,164 words across 2,242 segments, with the original timestamps preserved so you can click any line to jump to that moment in the embedded player.

What you can do with it

Use the transcript to take notes, quote the speaker, build a study guide, generate a summary with ChatGPT or Claude via the YouTube Summary tool, or export it as a timed subtitle file with YouTube to SRT. You can also re-open it in the transcriber to choose a caption track available on the video.

Free YouTube transcript tool

YouTube2Text is a free YouTube transcript generator — no signup, no daily limit. Paste any YouTube link and get the full transcript instantly, with timestamps, click-to-jump, caption language selection, AI prompts for ChatGPT, Claude, and Gemini, and exports to TXT, SRT, VTT, or Markdown.