YT YouTube2Text

YouTube2Text → Transcripts → Lecture 4: Structural transformation, home production, and labour markets

Lecture 4: Structural transformation, home production, and labour markets — Transcript

by CEPR & VideoVox Economics · 13,843 words · 2,000 segments · language en · Watch on YouTube

Open in transcriber Summarize with AI Download SRT/VTT

Full transcript

0:00for
0:49okay
0:50welcome um to this
0:53uh uh lecture um it's a part of the um
0:59cpr's steg program which stands for
1:01structural transformation and economic
1:03growth uh we want to thank uh the UK's
1:07foreign wealth um foreign Commonwealth
1:10and development office uh for funding
1:13this as part of the UK Aid effort um
1:17this is the fifth lecture in the class
1:20uh the previous lectures both the slides
1:23and the um recordings recordings are
1:25available on YouTube the both the slides
1:28and the recordings are available on our
1:30website uh today we're very happy to
1:33have uh Rachel and guy from
1:36lsse um who will be uh uh talking about
1:42uh home production labor markets and
1:44structural
1:45transformation um the those of you who
1:49have been here before know the system if
1:51you're a panelist if you're in the front
1:53row of the class you can unmute to ask
1:56questions um Rachel has requested
1:59clarification
2:00questions only clarification questions
2:02during the talk uh and then we'll have
2:05time for a broader discussion the last
2:0715 20 minutes at the end uh Doug gin and
2:11I will be in the Q&A for those who are
2:14uh just watching as a webinar and we'll
2:16try to answer your questions as best we
2:18can of course Rachel's more of an expert
2:20on the topic than we are
2:23um let me think uh oh those in the front
2:26row if you're not asking a question
2:28please mute your mic and I think that I
2:31covered everything um except uh on
2:34Thursday so this is the second lecture
2:36in a module on structural transformation
2:39last week be told herend dorf gave an
2:41overview of models um and but this is
2:45segueing into uh labor markets and so
2:48next week
2:50Thursday um Mark Rosen from Yale will be
2:53giving a supplemental lecture uh he's
2:55more of a micro development Economist
2:57but talking about some of the micr
2:58frictions in labor markets in developing
3:01countries um with that let me hand it
3:04over to Rachel go
3:07ahead thank you very much uh for having
3:10me and thanks a lot for Joe and duck to
3:13start this step program it um I'm sure
3:16it's going to be very useful for many
3:19people and many country and especially
3:21this series of courses um I think will
3:23be very useful for H graduate student in
3:26economics so today um I'm going to
3:30talking about structural transformation
3:31home production and labor market so as
3:34Joe already point out last week you had
3:38the a lecture by Berto on structural
3:41transformation so and that is the theme
3:44of this program as well so structure
3:46transformation as you know is about the
3:48changing structure in the economy in
3:50terms of both the output and employment
3:52share so typically as you seen last week
3:56that we saw the size of Agriculture
3:58sector uh is falling both in terms of
4:02employment share and the value added
4:03share and then we typically see a b a b
4:07shape or home shape in manufacturing and
4:10services is rising so that pattern is
4:12very robust when you look across many
4:15country and also look through uh the
4:17whole time series of the historical
4:20statistic for many developed country as
4:22well so today what we are going to focus
4:25on is another very important production
4:28sector and that is the home production
4:31sector and the Home Productions output
4:34is excluded from the GTP and the input
4:36is excluded from employment but it's a
4:39very important sector as I will show in
4:42showing you some of the measurement so
4:44what we are going to focus on today is
4:47what is home production how is it
4:48measured and why is it important and how
4:52is it related to structure
4:53transformation there are many many
4:55relationship but today because of the
4:57time I will focus on the implication for
4:59labor market and I will take two
5:02application one is the implication for
5:04the aggregate labor market and the other
5:06one is the implication for the labor
5:08market outcome by
5:10gender now I won't be going through the
5:12detail of the model or proofs um as once
5:16you learn the technique of how to solve
5:19this kind of structur transformation
5:21model from last week's lecture many of
5:23these model can be solved in a very
5:25similar way so I will not go through the
5:27proof and also you should refer to the
5:30papers to see the complete literature
5:32because I couldn't put all of them in
5:34the slides okay so what is home
5:37production um this is a book written by
5:41Economist Margaret Reed uh the books is
5:44really about the economics of household
5:47production what she defined as home
5:49production if you look at this quote
5:51there's few element first of all is the
5:55time spent at home that you're producing
5:58services and good for the member of the
6:01household okay and this output okay this
6:05time could potentially be marketized it
6:09can be done by a pay worker so someone
6:12else could have produced these goods for
6:15you and you just enjoy the output okay
6:18so for example you can think of cleaning
6:21okay instead of doing my own cleaning I
6:23could have hired a cleaning lady who
6:25cleaned the house for me and I enjoy a
6:28clean house so one key thing about home
6:31production and to make it SE different
6:33than Leisure is that we enjoy the
6:36product of Home production not the time
6:40spend on it okay whereas Leisure is
6:43something different Leisure is the time
6:45we spend on it that we enjoy so in other
6:48words I I will not hire someone to watch
6:51TV for me because that will not give me
6:53the utility of the Leisure I have to
6:56watch the TV myself whereas when it come
6:58to home uction I can hire someone to do
7:01the input for me and I will enjoy the
7:05output okay so this is the broad idea of
7:08Home production now of course home
7:10production is a concept that has been
7:11used quite extensively both in micro and
7:14macro there's some uh survey uh about
7:18the study that has been done so today we
7:20will focus mainly in the dimension of
7:22structure transformation and labor
7:24market so how do we measure home
7:27production okay so when we think about
7:30intuitively is basically the time I
7:32spend on it okay uh so you will turn
7:35naturally to some way of measuring how
7:38people spend their time so the first uh
7:42kind of well-known time use survey is
7:45the digitation by Le in
7:471917 where they recorded the time used
7:50for 60 family in Pennsylvania between
7:541917 until the first national Tech time
7:58you survey a available in the United
8:00States there various Time new survey
8:04they used to have a bureau of home
8:06economics which is kind of like the
8:08Bureau of Labor statistic but that
8:11Bureau used to basically have time you
8:14study they have station in different
8:16places where they collect you know the
8:19time budget of the housewife because
8:22back then not many women participate in
8:25the labor market so there's a great
8:27interest to know what they are doing if
8:29they're not participating in the labor
8:32market so it's also some way later on
8:35for people to think about the extended
8:37national account because these are
8:39resources that's put into production but
8:42doesn't recorded in the GDP accounting
8:45the national accounting so the first
8:47national survey started in
8:491965 and then you have a Sero wave of
8:53time survey and then by about um 2003
8:57you have this uh American time new
8:59survey in the beginning it was uh every
9:02two years then become every year then it
9:04change again after the Great Recession
9:07when the budget is a bit limited so
9:09there's a some academic paper discussing
9:12this time survey the two good example is
9:14aru and her and Rian Frances now of
9:17course there's also time use for many
9:19other
9:20country if you want to do do comparison
9:23across country the two well-known one
9:26are the harmonized European Time use
9:28survey which for mainly the European
9:30country and then you have the
9:32multinational time you survey which
9:34cover a brad the sets of country and of
9:36course there are many time you survey at
9:38the national level in many country so um
9:42if you look at the publication of the io
9:45by the way I have all these refences at
9:47the end of the lecture slide so you can
9:49refer to them uh when you get a copy of
9:51the lecture slide online so I I has a
9:54recent book that basically summarize all
9:58these uh home production ity and has a
10:00summary of 64 National time you survey
10:03and there is also effort in African
10:06country to have more time use survey
10:09because the gender gap in the time use
10:11is included in the African gender and
10:14development index okay so you can refer
10:16to this paper for that just to give you
10:18a sense what it is like uh about what
10:21people do in home
10:23production so um in your own personal
10:26experience you might realize your home
10:29production as a student very often may
10:31be including fixing a lunch for yourself
10:34cleaning uh your bat you know so it will
10:37be like cooking and cleaning mainly and
10:40indeed these are the major activity even
10:42for housewife for the last about 100
10:45years in the US you see that the cooking
10:48and cleaning is quite a big chunk of the
10:51time used and also when you look at um
10:55the developing countries such as here
10:57you have Ghana and South a Africa
10:59example you see that cooking is quite a
11:02big number okay and that number is quite
11:05similar to the time when the US in the
11:081920 and of course you can think of many
11:10reason why cooking time has DEC over
11:13time it could be because now in America
11:16people will go and buy more fast food
11:19okay there more cheaper fast food around
11:22for them to Outsource their home
11:23production time it could also be there
11:26are Innovation that make cooking easier
11:29such as the oven you know all the
11:32household durables okay all this could
11:35be the reason so when you look at the
11:38how time use at home has changed there
11:41is really like as if there's a
11:43structural transformation happening at
11:45home that family American family for
11:49example is spending less time on cooking
11:52and cleaning but more time on caring and
11:55household management so household
11:57management would be like doing some
12:00accounting uh of your of your household
12:03budget or you know uh shopping and
12:06things like that so these are the in
12:07terms of household management so there
12:10is a reallocation away from some of the
12:13thing that now people usually buy from
12:16outside into more doing the things that
12:19they do not Outsource as much and there
12:21could be also other reason that the
12:23total amount of this homegood might be
12:26changing over time as well so in the
12:29past maybe people do not have put as
12:34much attention to cleaniness okay and
12:37they might you know now care more about
12:40cleaniness because of the germ theory
12:43that show that it's important to have a
12:45clean home so even with the washing
12:47machine being discovered we do not see
12:50the cleaning time or laundry time
12:53reducing by too much we still spend time
12:55during the laundry B because maybe we're
12:57doing it more often we cleaning the
12:59house more often okay so those are all
13:02the factor that is behind this possible
13:05time pattern but there is this General
13:08pattern and in the developing country
13:11there are other item that are important
13:14for example they do spend a lot of time
13:16in know which part of the cooking is
13:18like collecting water and wood for
13:21example so these are the some components
13:23specifically in the developing country
13:26there have been some attempt to include
13:28the value of Home production to create
13:30an extended GDP because the purpose for
13:34us to measure GDP really is to
13:38understand you know the economic
13:39activity in the country and also to have
13:42a sense of welfare how many goods and
13:45services is out there so home production
13:48as you can see is quite a big chunk of
13:51time it's like you know 50 hours per
13:53week is more than a full-time job so if
13:55we included the employment you know the
13:58market employment in the GDP then this
14:0150 hour could potentially be producing
14:05quite a lot of goods and services in
14:07comparable to the GDP number so this is
14:10why you know there are effort trying to
14:12extend uh the national account and to
14:15see how we can measure the value of Home
14:17production in general there are two way
14:20you can do it you can think of the
14:21output evaluation so meaning that I
14:24compare what's the value of cooking I
14:26can look at you know if I have to buy
14:28this
14:29uh plates of spaghetti you know this
14:32spaghetti Bond only s if I go to
14:34restaurant how much will I pay and now
14:36at home I produce it so you can use that
14:39as a value you can also use a input
14:42valuation approach which is basically to
14:45put the value on the input such as the
14:48time you use will be based on the wages
14:51and then the household youo would be
14:53like to return to househ way household
14:55durable so this is precisely the method
14:58used by Ben Bridgman in his finding for
15:02the United States so he tried to compute
15:05the extended GDP for the United State by
15:08including the home production and he
15:10find the number to be very substantial
15:13so if you include the GDP from the home
15:16production then the extended GDP is 42%
15:20higher than the measure GDP so it's
15:24quite a big number and this number
15:26decline over time PR main because we see
15:30that home production time do decline in
15:32the United States still is quite a big
15:34number in 2010 it's still
15:3727% so this I hope to convince you home
15:41production is an important production
15:43sector both in terms of the time spend
15:46there and in terms of the potential
15:50value that It produced okay and this is
15:53why we wanted to spend some time
15:55thinking about it and then when you look
15:57at how home production change and we've
15:59seen from that time news table we did
16:02see there's a lot of reallocation so
16:04Margaret Reed writes in
16:061934 and talk about how the we are at
16:11there's a transformation within home she
16:14emphasized one form of production after
16:16another have in part transferred to
16:20commercial production okay and then even
16:24in 1930 she already observed Child Care
16:27education in care of the sick and are
16:30now to a large extent carried by pay
16:33worker so back then you know in 19th
16:36century education is usually you know
16:38done at home and the school is really a
16:41very recent phenomenon so that is one of
16:44the example so and then 1973 vanet has
16:48this P disertation so her supervisor is
16:52the Pioneer who started the uh time use
16:56survey the national time use survey in
16:58the United state so she really look into
17:01the first wave of the national time
17:03survey and then she also document the
17:06same kind of reallocation of task within
17:09the home production again as we've seen
17:12from the table before the household
17:14management part is getting bigger and
17:16bigger in terms of the time span there
17:19Leb go look at this at a more you know
17:22clear angle basically by looking at
17:24consumption expenditure so what he
17:27document is look a lot of these home
17:30production which is cooking and cleaning
17:33now is reflected in consumption
17:35expenditure the household are buying
17:38some of these product so his view is a
17:42lot of the you know housewi has switched
17:45from being a producer to a consumer
17:48because he directly look at the
17:50consumption expenditure and to document
17:53that a lot of this has been you know so
17:56So-Cal what we say Market so this kind
18:00of process of marketization of Home
18:02production meaning that over time you
18:05spend less and less time doing this home
18:07production yourself instead you're
18:09paying someone else to do it it will
18:11show up in the uh expenditure
18:14consumption expenditure it will also
18:16show up as the employment structure of
18:20the economy we will see sector that
18:23produce substitute to the home
18:25production expanding and this sector
18:28very often are the service sectors okay
18:31so some reent work by Freeman and Shak
18:34really you know push this thoughts of
18:37marketization strongly by using some
18:40cross country at time use data to
18:43highlight when you compare differences
18:45in hours of work across country is not
18:48the differences in the total hour but
18:51it's the differences in how many of
18:53these total hour are at home and how
18:56many of these hours are in the market
18:58Market okay so that is a quite an
19:01interesting line of study there so let
19:05me now turn to my first application
19:08which is look at the interaction between
19:10structure transformation home production
19:13for the agregate labor market so this
19:16application focus on one important
19:19variable which is hours of work so for
19:22many years growth and business cycle
19:25model in macro very often they talk
19:27about this concept of
19:29balance the balance growth path is
19:32referring to the scenario that the
19:34aggregate variable grow as some constant
19:37rate now constant rate include zero so
19:41in that kind of balanced growth path
19:44very often they will focus on the
19:46model's prediction that ours work is
19:49constant okay so they will have a
19:52constant hour work and constant hour in
19:55Leisure okay so that is a typical kind
19:58of model for macro and you can find it
20:01in the very um classical bat G business
20:04cycle by Kul and Prescot and for that
20:07King proa and reell had develop this
20:10balance growth preferences so this
20:12balance growth preferences basically
20:15will imply in equilibrium ours work will
20:18be constant but recently we see more
20:21work looking at the data that shows that
20:25market hours per capita are not constant
20:29okay ran Francis has the work looking at
20:32the US historical data for the 20th
20:36century and find that ours work um first
20:39fall and then mildly increase later on
20:43there's also a lot of cross country
20:45evidence about hours of work being not
20:48constant so Rogerson talk about how in
20:51some country the decline never come back
20:54and in some other countes such as us
20:56ours Decline and then come back that
20:58later on so the reason Work by Big
21:01funder and Lagos they have a biggest
21:05sample size of country look into more
21:08country at different development stage
21:10and you see similar picture there
21:12there's a big four in ours work during
21:15the early development stage and at later
21:17development stage some of the country
21:19has increasing and some do not
21:22increase so with this changing Market
21:26hours per capita that's sort of put the
21:28challenge to the macro model that use
21:31the traditional balance growth
21:33preferences where our words stay
21:35constant so a recent paper in the JP by
21:39B and crel they basically proposed a
21:42generalized balance growth part
21:44preferences basically generalize the K
21:47proc Rebel occurrences and with that
21:50preferences they can obtain a changing
21:53hours of work along the balance growth
21:55path okay so then you H you could have
21:58the aggregate balance growth consistent
22:01with hours changing and that is quite an
22:03important feature especially when you
22:05look at the US data because we see that
22:08the US growth has been pretty much at 2%
22:12since 1870 but the US hours of work were
22:16not constant so that is one approach to
22:18it today I'm going to focus on a second
22:22approach which is to think of how I can
22:25get a balanced growth path and change
22:28ing Market hour by thinking carefully
22:32what is nonmarket hour okay so in this
22:36line of research they very often think
22:39of hours work is Market hour and hours
22:42you don't work is just non Market hour
22:45they do not differentiate what are
22:48inside the non Market hour as we seen
22:51earlier on in the time new study I show
22:53you inside that nor Market hour maybe
22:57half of it is home production and the
23:00other half is Leisure and in the data
23:04Leisure and half Market out at home hour
23:07has a very different Trend so today my
23:10focus is to look at if I explicitly
23:14account for the norm Market hour has
23:17both home hour and Leisure hour how can
23:20I use the story of structural
23:22transformation to generate a trend in
23:26Market hour along the balance growth
23:28path okay so that will be my first
23:30application so here's a picture from
23:32Francis and Ry and this picture is quite
23:34typical for many country that during the
23:37early development stage the average
23:40hours of work in the market is falling
23:43and in United States it come up toward
23:45the second half after the
23:481980 the model will be very similar to
23:52the model that you have studied um last
23:55week uh under structure transformation
23:58with a key important new component which
24:01is the home production I'm going to base
24:04the lecture on this paper but I'm going
24:06to use a simplified version so I would
24:08say that it's only home production for
24:12services so you will have utility
24:15consumer demand different kind of final
24:18good there are three types of final
24:20goods there could be agriculture
24:23manufacturing or
24:25services but man services can be
24:29produced either in the market or at
24:32home now the Leisure is one minus L okay
24:38where L is the total hours of work so L
24:42include both homework and Market work so
24:46this is one of the crucial difference
24:48compared to the standard macro
24:50literature where one minus L is simply
24:53treated as non-market work and it's just
24:56the same thing here we are going to
24:59break down this L into home hours and
25:02into the market hour okay we are going
25:05to look at them
25:07separate so how does the interaction
25:09happen and what is the important
25:11components of it one important thing as
25:15we've seen from many of these historical
25:17uh study and also from our own
25:20experience is that many of these home
25:23production could potentially be
25:25outsourced in other word it means that
25:28the market and Home Service are good
25:31substitute okay they're not perfect
25:34substitute okay the bolog I made Might
25:38taste different from the one that I buy
25:40from an Italian restaurant but they are
25:43good enough as a substitute they're not
25:45perfect but they are good enough meaning
25:47that if the price in the restaurant drop
25:51it might induce me to buy more from the
25:54restaurant okay so that is the good
25:57substitute assumption on the elasticity
25:59substitution between home and Market
26:01services so that Sigma is bigger than
26:04one Epsilon though is the substitution
26:08across different type of good across
26:11agriculture manufacturing and services
26:14so as you have KN learned from the last
26:16week lecture most study will claim that
26:19that substitution is quite substantially
26:22below one okay so they are not as good
26:25substitute meaning that as we become
26:29richer as I consume more food and have
26:33nicer clothes I also want better
26:35surfaces okay I want better car and
26:38better health care services so it's
26:40better food okay these things they are
26:43not very good substitute okay and then
26:47there will be a production function just
26:48to close the model the production
26:50function will be identical across sector
26:53except this productivity parameter okay
26:56so to highlight this machism through the
26:58changes in productivity that is being
27:01different across sector so gamma J will
27:03be the productivity growth and there
27:05will be four sector in this economy the
27:08agriculture sector the manufacturing
27:11sector the market service and the home
27:14surface okay so these will be the four
27:16SE so you can go ahead and solve this
27:20model in three stages okay think about
27:23first individual has to allocate time
27:26between home and Market and in that
27:29allocation decision it basically will
27:32have this implication of about how
27:35individual will marketize their home
27:38production once you figure out within
27:41the good how you are going to allocate
27:43between home and market then you look
27:45across different Goods how are you going
27:48to allocate time across agriculture
27:50manufacturing and services and that step
27:53two is exactly what you did last week in
27:56terms of the structural transformation
27:58across the different
28:00sector finally once you have solved all
28:03these static allocation you turn to the
28:06dynamic allocation and saying that now
28:09for the whole economy how do I derive
28:12the balance growth path by allocating
28:14Capital over time okay uh by deciding
28:19consumption and saving decision so that
28:21give you the capital accumulation so the
28:24mathematics is very straightforward the
28:27static allocation is about equating
28:29marginal rate of substitution to
28:31Marginal rate of technical substitution
28:34and when you look at the damate you
28:37basically try
28:38to use step one and two to transform the
28:42model into a one sector model and then
28:45that go back to your Ramy growth model
28:48and then you can derive the balance
28:50growth okay so skipping the math
28:53basically this is from the marginal rate
28:55of substitution equal to the margin rate
28:57of technical sub solution you derive a
28:59very intuitive formula meaning that the
29:02share of Labor that is allocated to the
29:05market for the services compared to the
29:08share of Labor allocating to home
29:12services that difference will depends on
29:16the differences in the productivity
29:18growth and how good substitute they are
29:22intuitively all is saying is that if
29:26these are good substitute my home C food
29:30and the market bought food they are good
29:32substitute with a sigma bigger than one
29:35then naturally when the market sector
29:38has a faster productivity growth than
29:41the home sector we will Outsource the
29:44home production I will buy more from the
29:47market instead of producing at home so
29:50this is what we call the
29:52marketization okay due to the fact that
29:55the market service sector has faster
29:57productivity grow growth than the home
29:59service sector then you can go to your
30:02step two which is to look at how do I
30:05allocates time across sector and that's
30:07the structure transformation using very
30:10similar mathematics you can derive the
30:13labor share into two different kind of
30:17goods is has a similar formula as that
30:20formula over there is essentially saying
30:24the same thing if the elasticity of
30:27substitution is less than one then I
30:29will reallocate to the sector that has a
30:32slower productivity growth
30:35intuitively if you know the car industry
30:39is having very fast productivity growth
30:42compared to the health industry I want
30:47good car I also want good health how can
30:50I achieve both I will have to realate
30:53labor away from the car making industry
30:56into the healthcare industry because
30:59that's the only way for me to get more
31:01Health Care Services okay so in this
31:04case you're reallocating resources from
31:06the fast growing productivity sector to
31:09the slower growing productivity sector
31:11because Epsilon is less than one now of
31:15course the sector itself especially for
31:17services in this formula will be a comp
31:21composition of the market service and
31:23the home service so what you see here
31:26the marketization Force itself will
31:29increase the productivity growth of
31:31services because you're are allocating
31:34more labor into the market compared to
31:37the home okay so these are the math
31:40behind it but the picture is very
31:43simple the first stage you decide
31:47whether you should produce it in the
31:48market or in home because Sigma is
31:51bigger than one and because Market tfp
31:54has a faster growth you reallocate labor
31:57from home to market so this is this one
32:00stage and you go to the second stage
32:03because Agriculture and Manufacturing
32:06has higher tfp growth than services and
32:09these three are poor substitute with
32:12Epsilon less than one you will have to
32:14re locate labor away from agricultural
32:18manufacturing into services so that will
32:21be the flow of Labor and you can
32:23aggregate them into one total
32:25consumption and then solve the more
32:27model like a Ramsey growth model to
32:30derive the balance
32:32growth so as I said we skip the math I
32:35focus on the implication so what are the
32:37implication for the labor market now
32:40what we see is home production hour
32:43subject to two forces the first forces
32:47is this U marketization Force because
32:51Market sector is growing faster so you
32:54will Outsource your home production so
32:56home hour will four due to
32:59marketization but home hour will also
33:01gain hours due to the structure
33:04transformation so in this diagram you
33:06can see that as resources are
33:10reallocating from agriculture
33:12manufacturing into Services part of the
33:15service is producing at home so there is
33:18a positive force of Labor coming into
33:21the home production okay and this is why
33:24structure transformation is pushing up
33:28the home production marketization is
33:31pushing down the home production there's
33:33these two forces at
33:35work now let's turn to the market hours
33:38especially on the agregate market hour
33:40that we were interested in in the
33:43beginning so what can we think about so
33:46there's these two forces there structure
33:47transformation and
33:49marketization which force is going to be
33:52stronger well in the beginning you know
33:55if you go back to the early you know uh
33:5920th century or the middle of the 19th
34:01century let's say United States over 50%
34:06of the labor force in 1870 is in
34:09agriculture so we have a very big
34:11agriculture sector so that means you
34:14will have a lot of Labor that is here
34:17once the structural transformation
34:19forces started because of the
34:21differences in tfp growth you see that's
34:24a strong force leaving agriculture to
34:27come into
34:28services that force is so strong it will
34:32dominates this marketization Force
34:35because perhaps the gap between the home
34:37and Market service tfp grow was not that
34:40big so this force will dominate this
34:43force and as a result we will see a rise
34:46in home hour and a fall in the total
34:50Market
34:51hour and at the same time in terms of
34:54the structure of the economy you will
34:56see a fall in agriculture Rising
34:58manufacturing and the service sh so that
35:01characterized the time dimension for the
35:04United State picture from 1870 to about
35:08maybe 1950 or 1940 so you will see a
35:13decline in Market hour Po in agriculture
35:16Rising services and Manufacturing but
35:20then by 1950 the agriculture employment
35:23is very small okay it's already you know
35:26almost like % very small so then you go
35:30back to this flow diagram What's
35:31Happening Now agriculture is already
35:34very small you know less than 10% so
35:37there's very weak forces coming out of
35:41Agriculture at this point there will be
35:44a much stronger forces between home and
35:46Market okay because the home surface has
35:49been growing so now is the time that is
35:51really losing a lot of Labor or
35:54Outsourcing a lot of Home hours into the
35:57the market service sector so in the
36:00second part of the story what we see is
36:02the marketization force now will
36:04dominate the structural transformation
36:07Force home hour will start to fall and
36:10you will see the rise in the total
36:12Market
36:13hour in terms of the structure of the
36:15economy because the agriculture sector
36:17is already so small manufacturing at
36:20that time will be about 30 to 40% that's
36:23the time when manufacturing is really
36:25Rank and then we see manufacturing
36:28slowly decline from the 40% to the 20%
36:31we see now and the services rise to
36:34about 80% so that give you the picture
36:38of what the model will predict when you
36:40look at the interaction of structural
36:42transformation and marketization for the
36:45hours in this economy and then when you
36:48look at the aggregate balance growth
36:50path as I said you can convert the model
36:53to essentially one sector solve the
36:55ransing model and then derive the
36:57equilibrium balance growth part what you
37:00will see is the total work is constant
37:05because we still have this balance
37:07growth path preferences but that
37:10preferences is defined over one minus L
37:13and L is total work okay so what is
37:18restricting is the total work is
37:20constant but it doesn't imply or require
37:24Market work is constant so that is very
37:27nice because because then you can
37:28capture what we see in the data market
37:30hours is actually
37:32falling and you have the balance growth
37:36part because the aggregate Capital stock
37:38and the market output will be growing at
37:41the constant ratees of the labor
37:42augmenting productivity growth in the
37:44manufacturing sector so that will be
37:47your underlying aggregate balance growth
37:49path on top of this there are Trends in
37:53Market hour which we have already
37:55described such as the falling AG culture
37:58the H Manufacturing and Rising services
38:01and most importantly we see that the
38:03Agate Market hour with first FL then
38:06rise even though the aggregate total
38:09work will be constant along this balance
38:12growth far so what this example show is
38:15it's very important to look at what is
38:19total work okay the total work composed
38:22of two components so when we have this
38:26balanc growth it doesn't necessarily
38:28require the market hour to be constant
38:31okay so that is one of the important
38:34insight to see once you take to into
38:37account what we observe in the data
38:40which is structural transformation and
38:42magnetization then you can also explain
38:44this nice um know this pattern in the
38:48data in terms of the aggregate Market
38:50hour well if you ignore these components
38:53you won't be able to explain them so
38:56that was the Benchmark model where total
38:59work is constant now there are some
39:02discussion especially about from Agri
39:05and her that shows that leure in the US
39:08has been rising since the
39:111965 at First National time use survey
39:15so they report quite a substantial rise
39:17in leure so how can we use this model to
39:21think about the transends in lure in The
39:24Benchmark model leure will be constant
39:26because to work is
39:28constant so when we were working on this
39:31and we were confront with this question
39:33so we started to think
39:36about why you know in the data is leure
39:39is rising but in the model leure is
39:42constant what are we missing and when we
39:45look at the way that is set up in the
39:47model we realize leisure in The
39:49Benchmark model is just this V depending
39:53on Leisure so it's the time itself that
39:57you
39:57utility okay so in the model in the
40:00Benchmark model leure is just utility of
40:04time your free time but if you look at
40:07modern day Leisure like when you finish
40:10this lecture when you want to enjoy
40:12yourself when you want to relax you will
40:15not be just sitting there you know
40:17relaxing you know enjoying the time per
40:20se very often modern day leure
40:24involve combining your time with capital
40:28goods such as you might take out your
40:31phone start the browsing on your phone
40:34you might turn on your computer you
40:37might turn on your TV you might turn on
40:40your video game all these kind of com uh
40:43Capital component or put it more
40:46broadly you might be enjoying it with
40:49some uh nice furniture around your house
40:52so all this is telling you that it's not
40:55time per se that generate utility from
40:59Leisure mod than day Leisure is no
41:02longer just about reading books and
41:04seeing your friends mod than day leure
41:07is about you know enjoying your time
41:10with some of these Capital GRS okay and
41:13so you can modify the model to do that
41:16once we introduce the food model where
41:19you allow the production of leisure good
41:22so you enjoy the time and also you have
41:25to combine your time with some capital
41:27good
41:28then you can produce the results that
41:31there will be rise in leure that is to
41:34do with the production of leisure grip
41:37even though the traditional Leisure is
41:39still concern traditional Leisure I mean
41:42you enjoy the time only so the model can
41:45generate a increase in total leure and
41:48because of this increase in total leure
41:51quantitatively the model will be able to
41:53predict a bigger fall in the market hour
41:56in the beginning and a slower rise in
41:59the market hour at the end so you will
42:01get more instead of very you know you
42:04like this you will get more like an L
42:07which is closer to what we see in the
42:09data so more recently I have some work
42:12with Tima that look at these different
42:15Trends in leure across uh different
42:17skill group as what we see in the data
42:20also document by Agri and HST is that
42:23the rise in Leisure is very different
42:25across group we see that low skill
42:27worker has a bigger rise in L compared
42:30to high skill worker so you can go to
42:33that so because of time uh I I'm going
42:36to just make this one more remark and
42:38then skip to the
42:39next so the second remark I want to make
42:42about is in terms of the theory so last
42:46week you learned with Berto about the
42:49two sources of structure transformation
42:52so that's how usually people in the
42:54literature talk about that they talk
42:56about income effect
42:57so they say that oh when income grow we
43:00see shift of the allocation resources to
43:03Services because demand for services is
43:07more income elastic than demand for
43:09goods and very often they site the angle
43:12curve meaning when we look at
43:15expenditure when household become richer
43:18we see the expenditure on Services
43:22increase by more okay so share of
43:25expenditure onto um Services increase by
43:28more and then there's a second source
43:31which is sometime called the relative
43:32price effect uh which is basically
43:35saying there's a change in relative
43:37price and because the elatic
43:39substitution is less than one so when
43:43the price of services go up you will
43:45have to spend more on Services sometime
43:48this is called the Bal
43:50effect so in the model I have just
43:53presented to you so usually when people
43:55do the income effect they think about
43:57non homematic preferences in the sense
43:59that the income elasticity for services
44:02is bigger than one the model we have
44:04just study preferences is homop fatic
44:08but what we have seen is the presence of
44:11Home production actually allow this kind
44:14of homopathic preferences to capture the
44:17essence of angle curve the idea is if
44:21home production is a better substitute
44:22to Services than good as we've seen
44:25there then when you have
44:27marketization that would be like an
44:30increase in the relative demand of
44:32services so even though everything is
44:35driven by the relative price due to
44:37productivity growth you can actually
44:40induce a income effect you will see that
44:43the expenditure on Services Will Rise by
44:46more simply because home production is a
44:49better C to to the services than good so
44:52there's some recent paper in the AJ Mel
44:55trying to capture how how important is
44:58this role played by the Home production
45:00in explaining that non-h homeopatic term
45:03for the services in preferences okay so
45:06and I will refer you to that I say
45:08before of time I'll be very brief here
45:11so there's a big literature about
45:13comparing hours of work across country
45:16so Rogerson has a very nice paper that
45:18highlight about the role of structural
45:20transformation and the low Market hour
45:23in Europe is mainly due to this uh small
45:26service sector in Europe and there are
45:29you know a lot related to that people
45:32then started to asking but what about
45:34Scandinavian country which has high tax
45:37so here in the Rogerson paper High tax
45:40would discourage marketization because
45:42in order for you to marketize you like
45:44paying two tax you first you go to work
45:47you pay tax you hire a nanny you also
45:51the pay the tax for the Nan so this tax
45:53is really increasing the cost of out
45:57sourcing your child care so in that
45:59sense you know there will be less
46:01marketization so the outline and people
46:03talking about Scandinavian country they
46:05will say well in Scandinavian country
46:08you also have very high tax but they
46:10have very high Market hour so there's a
46:12line of research there to look at the
46:14subsidy policy of the SC Scandinavian
46:17country so not spend much time we can
46:20come back to that later on in the Q&A
46:23session so the last uh 25 minutes I will
46:27talk about this second application which
46:30is how you know structure transformation
46:33home production is linked to the labor
46:36market outcome of uh gender by gender
46:40okay so this is clear H link to the
46:43recent literature on the rise in female
46:46employment so this literature is quite
46:49recent because um for many uh developed
46:52country of since around the 1960 1970 we
46:57see female employment
46:59Rising so a lot of work where uh
47:02economic research going to try to
47:04understand what other reason for that
47:06rise most of these work focus on gender
47:10specific Factor okay meaning that it
47:13could be a woman acquiring more
47:15education the human capital accumulation
47:18it could be that it's culture change uh
47:20social Norm are changing there could be
47:23medical advancement okay that is the
47:25pill the contractor pill and uh you know
47:28uh formula milk all this can contribute
47:31to the rise of female and they focus on
47:34this gender specific Factor now the
47:36gender specific Factor will basically
47:39imply a rise in female in intensity
47:42across the whole economy okay all the
47:46industry and this are mainly supply side
47:49mechanism is that woman coming into the
47:51labor market so sometime they will have
47:54a bit of difficulty to jointly account
47:57for the rise in female employment and
48:00also the rise in female wages because
48:02intuitively if you think of uh the
48:05household durable make women have more
48:08time so they decided to join the labor
48:10force that would be like a supply effect
48:13and that should sort of depress the
48:15female wage relative to male so it will
48:17have a difficult time to get a
48:19simultaneous rise in the female wages so
48:23in the recent work with barara pachango
48:25what we try to do is to propose a gender
48:28neutral mechanism that focus on
48:31structural
48:32transformation and trying to use that to
48:35explain why we will see both female
48:37employment and female wage
48:40increase now this J Mutual mechanism
48:44that the focus on structural
48:45transformation is not you know you know
48:49it is like a story told by many people
48:52okay so when you look at the evolution
48:55of Home production and structure
48:56transformation
48:57very naturally you see that there's a
48:59very important angle by gender in fact
49:02the time use table I showed you before
49:04was based on housewife okay
49:07woman so if you look at that ILO report
49:11they will tell you that based on the 64
49:13National time year survey women's home
49:16production hour a triple of men so it's
49:19women who are doing the home production
49:22name and very often when they ask woman
49:26why don't you participate in the labor
49:28market and they will say the main reason
49:30is because if I go to work who is going
49:33to take care of the household who's
49:35going to do the home production so home
49:37production become the main self reporter
49:39barrier for man the reporter barrier
49:42usually is about their health and other
49:45reason but for women the biggest item
49:48that women report is home
49:51production so when you think about that
49:53then the development of the market
49:55Services is very important because it
49:58allows women to Outsource the home
50:00production so that we Lee women to go to
50:03work that create one job and they also
50:06create a second job because the sector
50:09that often produce the substitute for
50:11home production are very often female
50:14intensive so if you think about the
50:16child care sector uh they will be mainly
50:19you know hiring a woman as well so this
50:23marketization process itself potentially
50:26who create two jobs for one okay and
50:29that is very important quantitatively so
50:32here is just some
50:34data um to show you the size of the uh
50:38home production hour by woman which we
50:41kind of show you in the beginning when I
50:43presented that table you know I said
50:45it's like 50 hour which is like a
50:47full-time job another thing to take away
50:50from this because now we can see more
50:52country and you can see the low and
50:56middle income country the oecd country
50:59the rich one and the Middle inome
51:02Country here so there seems to be a
51:06pattern going on here that along the
51:09stage of development home production
51:11hour for women are falling and that says
51:15something about what to expect in terms
51:18of M woman's time allocation how do they
51:21use their time and this is very
51:23important because across the globe we
51:26see that
51:27women's education is rising fertility is
51:30falling so there's a lot of talented you
51:33know educated women they you know are
51:36not uh having so many children as before
51:39so they have the time they could
51:42potentially participate in the market so
51:46understanding the time allocation
51:48between home and Market is very
51:49important for understanding the
51:51allocation of talent especially we see
51:54education of women has really converge
51:56to the man across the globe okay so that
52:01is one important motivation why we want
52:03to turn to that and historically and by
52:06many PE uh you know previous uh
52:09Economist sociologist they have
52:11emphasized the important link between
52:14female Market work and the rise in
52:17services and other also have emphasize
52:20this link about even within Services
52:24when you look at occupation if you think
52:27about some brain intensive occupation
52:29versus brwn intensive occupation that
52:32can act as some gender buyers demand
52:35shift and if Services is more brain
52:38intensive than Brawn intensive if you
52:41think about manufacturing is more BR
52:43intensive where surfac is more brain
52:45intensive and that itself can tell you
52:48how surfaces might link to the rise in
52:51the female Market world so in this
52:54particular paper what we look at is to
52:56say that if for whatever reason woman
53:00has comparative advantage in Services
53:03whether it's due to the brain versus BR
53:05or due to communication skill or face-o
53:08face interaction you know there are
53:10study that show you know for example in
53:13medical journal they show that patient
53:15tend to trust female doctor because the
53:18female doctor communicate better to the
53:20patient okay for any of these reason
53:23suppose woman have a competive advantage
53:26inut proding Services then what we find
53:29is the expansion of the service sector
53:31will create job for women and the
53:33marketization of Home production would
53:35draw woman work into the market and then
53:38together it's like this two job being
53:40created story I mentioned earlier on
53:43that will account for the rise in
53:45women's Market hour the four in men's
53:48Market hour and the four in the gender
53:51wage G because this is like a gender
53:54endogenous gender buyers demand shift
53:57okay even though the sh itself is at the
54:00sector level what it transform to be it
54:04become gender bias because of the
54:06comparative so I'm going to show you
54:08just some data for the United States so
54:12here is some uh data for men and women
54:15age 21 to
54:1665 um on this side you see the weekly
54:20hours okay so this is total weekly hour
54:23the whole of United States as we seen
54:25earlier on that's a g a small rise in
54:28the market hour for but this black line
54:32This agregate Market hour which FOC we
54:35focus on in our first application is
54:38very mild because it mask the very
54:41different Trend across men and women so
54:44over time Market hour for men is still
54:47falling from 1970 on continue falling
54:51from their previous Century Trend where
54:54for woman is different
54:56Market hour for woman has actually been
54:59rising during this period okay and when
55:02you aggregate them you get a kind of
55:05very mild pattern and then we have the
55:07structure transformation picture which
55:09is the service share in the agregate
55:12economy has been rising and it's also
55:14Rising for both men and women so both
55:17men and wom is working more in services
55:20but looking at the share is hiding
55:22something behind so what we do is let's
55:25look at the hour hours they allocate to
55:28it so the aate man hour was falling but
55:32where did they fall what this show is
55:35the fall in men alter is mainly due to
55:38the fall in the gr sector so men now
55:42work Less in the manufacturing sector
55:45for
55:46examp and their hours in Services
55:49actually didn't increase that much the
55:52share increase but the hours doesn't so
55:55all the four in the men's Market hour is
55:58due to the fall in men's hour in the
56:00goods sector the other side of the
56:03picture is the rise in female hour
56:06almost entirely come from the rise in
56:09females service hour okay so that is
56:12something quite important and here's a
56:14picture on the gender wage ratio you see
56:18that that's in all sector and for the AG
56:21economy the sh H wage ratio between
56:25female wage to now it increased from 65%
56:28to about
56:3175% and then you look at what about the
56:34time use of men and woman so here you
56:37see a very um interesting pattern over
56:41time woman is doing less home production
56:45okay so you see the woman's home
56:46production time decreased substantially
56:49from 965 on work and that was matched by
56:52a increase in the uh Market hour the
56:56other side of the coin is men's Market
56:59hour was falling luckily men is actually
57:02helping out at home so overall when you
57:05look at total hour it didn't change very
57:08much okay so there's something called
57:10the iso work meaning if you look at the
57:13total work between market and home
57:16there's not much difference across
57:17gender the difference is how they
57:20allocate this Market work okay over time
57:23women allocate more and more hours into
57:25the market okay West man is allocating
57:28less so what we are going to do is just
57:31to see how this model of structure
57:34change and home production can generate
57:37this kind of pattern so it's very
57:39similar to previous model in terms of
57:42the utility structure you still have
57:44structure transformation across two
57:46goods and the service good is produced
57:49in the market and in the home here's a
57:52production function now the main
57:54difference is we are going to think of
57:57female labor input and male labor input
58:01with a CES structure okay so both sector
58:05or all the production sector use male
58:07and female labor input but some sector
58:12use female more intensively than other
58:15so naturally you will think of the
58:17service sector and the home production
58:19sector is more female intensive than the
58:22good sector so these are the Assumption
58:25on the comparative advantage of female
58:28in producing
58:29ofer
58:31so how do we get to this gender wage Gap
58:34as we mentioned earlier which is a
58:37challenge for the supply Side
58:39Story think about a one sector model in
58:42a one sector model if you have a CES
58:45production function with male and female
58:47labor when you equate the relative wage
58:51to the marginal rate of technical
58:53substitution this is what you will
58:54obtain w AG will depends on their
58:58usefulness in the production function
59:00female versus male and inverse of the
59:05hours they work because of the
59:06diminishing marginal productivity so
59:10when you look at this and you think
59:12about the data in the data we know LF
59:16over LM is rising over time so that
59:19means LM over LF will be falling over
59:22time so that rise in females and
59:26appointment should have a downward
59:28pressure on WF over WF so if you only
59:32look at the supply side mechanism that
59:36induce a rise in LF over LM you B to get
59:40a full in females wages so in order for
59:44you to get a rise in female wages you
59:47need this to be changing and sometime
59:49they call this gender bias demand shift
59:52okay so very often you you have to make
59:55some assumption you can say oh because
59:57of the fall in
59:59discrimination this term is rising over
1:00:02time okay that could be the story now in
1:00:05the two sector model though this rise
1:00:08will occur naturally and the reason is
1:00:12when you have a two sector model the
1:00:15relative wage has to equalize across
1:00:17sector and each sector will have this
1:00:20relative usefulness and the
1:00:23intensity if there is structure
1:00:26transformation you move from less female
1:00:30intensive sector to more female
1:00:32intensive sector it's like you're moving
1:00:35labor away from the side g into side
1:00:39s that endogenously will imply a
1:00:43increase in the aggregate side and this
1:00:46is what I mean by the structure
1:00:49transformation Force which is similarly
1:00:52General neutral because it's driven by
1:00:55some sectorial tfp growth that induce we
1:00:58move from G to S and endogenously imply
1:01:03the aggregate side to go up so it acts
1:01:06like a genda bias demand shift but it's
1:01:10driven by a sector bias um reallocation
1:01:15okay so this is how a structure
1:01:18transformation model can give you
1:01:20endogenous rise in female to male wage
1:01:24ratio okay now in terms of the hours you
1:01:27can use the story we learned about
1:01:30previously on the Agate Market hour so
1:01:33it's the same story here structure
1:01:35transformation will decrease Market hour
1:01:37for everybody because remember that flow
1:01:39diagram I was showing you structure
1:01:41transformation is bringing some hour
1:01:43into home production marketization will
1:01:46increase hours uh Market hour for all
1:01:49but because woman has this comparative
1:01:52advantage in services so that means the
1:01:54marketization
1:01:56would bite more for women so this source
1:01:59of increasing Market hour will be much
1:02:02stronger for women than for men and this
1:02:04is how you can predict a rise in women's
1:02:07Market hour relative to men's hour hour
1:02:10so overall what we see is a similarly
1:02:14gender neutral mechanism through sector
1:02:16specific productivity growth imply a
1:02:19rise in the service sector okay and that
1:02:22rise in service sector is both a
1:02:24combination of structure transformation
1:02:27and
1:02:28marketization and that together will
1:02:30imply a increase in the females relative
1:02:33employment and relative wage okay so the
1:02:36model uh we calibrate the model and we
1:02:39show that this is quantitatively
1:02:41important and we show that this kind of
1:02:43structureal transformation force can
1:02:45explain quite a substantial variation in
1:02:48the time allocation now I just wanted to
1:02:51come through the last few slide and then
1:02:53I will start for the Q&A
1:02:56so far even though I have mentioned
1:02:59about this intuitive story about
1:03:01household doable for example so far I've
1:03:03been keep emphasizing the role of
1:03:06marketization meaning that the only way
1:03:09women can reduce their home production
1:03:12hour is by Outsourcing so there has been
1:03:15um a very um uh popular paper called the
1:03:18engine of liberalization by uh Greenwood
1:03:21society and yor grouth and they
1:03:25highlight the importance of household
1:03:28durable so they are saying that if
1:03:30household durable become cheaper
1:03:32household can now substitute the time
1:03:35with the washing machine with uh the
1:03:39oven and vacuum cleaner all these things
1:03:43so how does this uh coming in well of
1:03:46course this has a role in fact there are
1:03:48some you know CrossCountry evidence uh
1:03:51among the oecd country to show the
1:03:53important of this but there is one
1:03:56important question that it might not
1:03:59have played such an important role
1:04:02historically is that there has been this
1:04:05something called Coen Paradox which is
1:04:08saying that if you know there is so the
1:04:12question is why did Homemaker end up
1:04:14working longer hours in their home in
1:04:16the century after 1870 despite the
1:04:19growing mechanization of household
1:04:21activity so then what they discover is
1:04:23despite as a diffusion of the this kind
1:04:26of household durable but uh you know
1:04:29housewi are spending more time cleaning
1:04:31and cooking and that is because of at
1:04:34this time as proposed by the historian J
1:04:37here is that this is also the time in
1:04:39the 1930s this is also the time when
1:04:42germ Theory Advance when the nutrition
1:04:45theory is becoming important so back
1:04:48then in the 19th century people don't
1:04:50eat hot meal because it's so time
1:04:52consuming to cook like hot meal but now
1:04:55with the availability of household
1:04:57durable is doable and people realize
1:05:00that it's important for nutrition value
1:05:03so they do more cooking now similarly
1:05:05they clean more so what happened is it
1:05:09actually they increase the demand for
1:05:12this kind of home production group so
1:05:14that is one reason that it could have
1:05:16created more homework and there was also
1:05:19another story that Insight point to that
1:05:22you know because of washing machine
1:05:24people do not bring their cloth to the
1:05:25laundry M instead they take them home to
1:05:28watch because there's some negative
1:05:30utility Associated using public Capital
1:05:33so Joe is one of the co-author here and
1:05:36then you know basically they argue that
1:05:38it could be a demarkation happening
1:05:41because of the availability of household
1:05:43durable so this is uh you know still
1:05:46there it's important but it's not the
1:05:48only dominant Factor marketization
1:05:51itself can be quite important another
1:05:54thing the final remark on on that is the
1:05:56rise in female employment is actually a
1:05:59rather recent phenomenon for sets of
1:06:02developed country if you look at the
1:06:04data in many country okay developing
1:06:08country now female employment is
1:06:11actually falling if you look at for
1:06:13example India and China okay they are
1:06:16growing they're getting richer but
1:06:18instead of rising female employment they
1:06:20actually experience falling female
1:06:23employment now this is not just
1:06:26because it's China or India if you look
1:06:28at the historical data for many odd
1:06:31count actually you do see the same
1:06:34pattern female employment was declining
1:06:36in the early stage of development stage
1:06:39sometime people call this a u shape
1:06:42pattern meaning during the early
1:06:44development stage female employment is
1:06:46falling and later on the female
1:06:48employment started Rising again so I
1:06:51have some recent work that try to you
1:06:54use the framework structureal
1:06:55transformation with home production and
1:06:58show that the importance of agriculture
1:07:01in explaining this kind of usate pattern
1:07:04okay and there is also other dimension
1:07:07when you want to think about structural
1:07:09transformation and labor market which is
1:07:12the dimension of skill so Joe here is
1:07:15one of the pioneer on with on this with
1:07:19Paca so in their paper in talking about
1:07:22the rise of the services they point out
1:07:24a very interesting and important fact
1:07:27which is that the RIS in services are
1:07:29concentrated in high schol Services okay
1:07:33so there H recently they also talk about
1:07:36how this skill bias structure sh
1:07:38contribute to the rise in the skill
1:07:40premium so I also have some recent work
1:07:42looking at how the rise in the relative
1:07:45price of high schol sector not only
1:07:47contribute to the rise in skill premium
1:07:50they also contribute to the stagnation
1:07:52in the low skill workers wage and the
1:07:55wage productivity Divergence now it's
1:07:57important to understand that riseing
1:08:00skill premium and low skill wage
1:08:02stagnation are related but one does not
1:08:05necessarily imply the other you can have
1:08:08both low skill wage and high skill wage
1:08:10Rising but just that high skill rage
1:08:13Rising faster the low skill rage station
1:08:16is really about the level of the low
1:08:18skill work and this kind of SK dimension
1:08:22of course related to some uh recent work
1:08:25by um pioneered by the David arter in
1:08:29his manyu paper uh in his recent paper
1:08:31he focused on structure transformation
1:08:34and
1:08:34polarization there when they talk about
1:08:37polarization they say that the wage and
1:08:39employment grow concentrated on the
1:08:42upper and bottom of occupation skill
1:08:45distribution now when you first look at
1:08:48this you thought wow is this
1:08:49contradicting the skill bias one now the
1:08:53answer is no because when they talking
1:08:55about occupation skill distribution that
1:08:59skill distribution is actually measured
1:09:01by the mean wage of the occupation and
1:09:04this is quite different because when we
1:09:07talk about low skill wage stagnation and
1:09:10riseing skill premium we're talking
1:09:12about low skill worker okay worker who
1:09:15doesn't have a college degree for
1:09:17example but
1:09:19polarization is about the skill of the
1:09:22occupation and the way they measure it
1:09:24for example in ut and Dawn they used the
1:09:27mean wages of the occupation in 1980 and
1:09:31that ranked the skill of the occupation
1:09:34and then look at what happened okay so
1:09:37these two phenomenon may be related but
1:09:40the relationship is quite subtle so
1:09:43recently there's a dissertation by ens
1:09:47which show that within an occupation
1:09:50there's a lot of skill hogen and the
1:09:53mapping between workers skill and the
1:09:56occupation has been changing over time
1:10:00and this explain why we see
1:10:03simultaneously polarization and also the
1:10:06rise in skill premium so these are
1:10:08something you know to the broader lature
1:10:11okay and the story of structure
1:10:13transformation are very often also
1:10:16involved in that literature on the
1:10:18occupation and I think my time is up I
1:10:21will stop here and I believe that was my
1:10:24last life and
1:10:26yeah so I'm very happy to have more
1:10:28discussion on the
1:10:35topic thank you Rachel do we have any
1:10:37questions especially from the front row
1:10:38you can just unmute and
1:10:41ask and people can enter them also into
1:10:44the
1:10:49Q&A uh yeah so can you hear me yeah okay
1:10:55I have a a question related to like the
1:10:58implications of including or not
1:11:01including home production into our
1:11:03measure of GDP so I was just thinking
1:11:06like um like it it seems like it's the
1:11:09case that in less developed countries um
1:11:12the share of Home production in
1:11:13aggregate output is uh greater than in
1:11:16more developed countries because more
1:11:18developed countries um there is more
1:11:21likely to Outsource these services to
1:11:23the market so I was just thinking if
1:11:25that's the case um is it also the case
1:11:29that we might be
1:11:30overstating
1:11:32CrossCountry um output differences and
1:11:36do you have a sense of like this may be
1:11:39like U an
1:11:40important something important or just a
1:11:43minor minor thing I I do think that is
1:11:46important in fact dck Goins has a very
1:11:49nice paper that show the importance of
1:11:53thinking about home production when you
1:11:54look at cross country income differences
1:11:57now it come in both way one is about
1:12:00measurement issue okay of course the
1:12:03facts that the uh de developing
1:12:05countries spend a lot of time in home
1:12:07production that tells you that
1:12:09especially if they spend more than the
1:12:11develop country I think it was that
1:12:13graph that I was showing later on I keep
1:12:16flipping the WR way so you're right so
1:12:18just from a measurement point of view by
1:12:20not including the home production we
1:12:23sort of exagerate
1:12:25the differences become the for the GDP
1:12:28because look the developing country they
1:12:31spend much more on home production if
1:12:33the output of these home production are
1:12:35not included in GDP then we sort of
1:12:37exaggerate the the welfare difference
1:12:40when we only look at the productivity
1:12:42difference or the GDP differences so
1:12:46that is one direct impact through the
1:12:48measurement and there's also you know
1:12:51interesting aspect in terms of you know
1:12:53the time allocation
1:12:55so I think that has a very nice paper
1:12:58that basically look at how in
1:13:00Agricultural Society you know because
1:13:02you know when they are on the Family
1:13:04Farm they could allocate the time you
1:13:07know H easily between home and market
1:13:10and so the input they actually put into
1:13:12the market production is not as much
1:13:15when you look at the developed country
1:13:17when you say oh they participate in the
1:13:19market they do fulltime so there are
1:13:22many Dimension this is going to matter
1:13:24quite a lot so just from a measurement
1:13:26point of view this is very important and
1:13:29there's also some interesting thing I
1:13:31discovered quite recently which is that
1:13:34um you have to be very careful that all
1:13:37the hours in home production um they're
1:13:39not in employment okay that we know for
1:13:43sure so what I said is there so if it's
1:13:45home production time it will not be in
1:13:47employment but the interesting thing I I
1:13:50look carefully into this day is that the
1:13:54output of Home production actually is
1:13:57not not all included in GDP okay so the
1:14:012008 sna when they trying to do national
1:14:04account they actually
1:14:07include the home production time for
1:14:11goods into the national account so home
1:14:16production is for own use but it could
1:14:19be production of services like all those
1:14:21example we talk about the dominant
1:14:23example cleaning laundry you know
1:14:25cooking but then in the early
1:14:28development stage there are important
1:14:30things such as making your own clothes
1:14:33uh fetching water and wood you know
1:14:35these so and also subsistent farming you
1:14:39know growing your own food this item
1:14:42this output is actually for reason I do
1:14:47not quite agree included in the GDP now
1:14:50in practice how they measure it is a
1:14:52different matter okay that could be a
1:14:54complete sleep Mass you know when we
1:14:56were looking at the African uh data you
1:14:58know some make a lot of effort to
1:15:00include the value from water and wood
1:15:02collecting some don't know what to do
1:15:04but in principle the the guideline is
1:15:08find if it's own used Goods is included
1:15:12in GDP but on new service is not
1:15:15included in GDP so when you look at the
1:15:18data you really have to be very careful
1:15:20exactly in that country did they really
1:15:22make an effort to do that but that is
1:15:24Mak the modeling very challenging
1:15:27because when we do our model we think of
1:15:29if the input is not in employment the
1:15:32output will not be in GDP but for this
1:15:35small category which is called the onuse
1:15:39goods they their input is not in
1:15:42employment but their output is in the
1:15:45GDP yeah so that you just have to be
1:15:47very careful when you look at specific
1:15:49of course it's not a problem for
1:15:51developed country okay because we don't
1:15:53do own use Goods anymore these they all
1:15:56home production or
1:16:00Services Rachel there's a question from
1:16:02the Q&A that I'll read to you it's
1:16:07Emanuel
1:16:08KY um is it correct to say that there is
1:16:11a rebirth of the Cohen Paradox due to
1:16:15covid-19 if so how does that also
1:16:17contradict the fact that in the current
1:16:19pandemic the majority of frontliners are
1:16:22women meaning women are both needed at
1:16:25the workplace and also at
1:16:28homes okay this Co is one of the main
1:16:32thing making you know the media also
1:16:36turn their attention to home production
1:16:38so if you look at the what's happening
1:16:41around the world in December all the job
1:16:45loss in United States is on women
1:16:49actually more than the total job loss so
1:16:52that means in December men is gaining
1:16:54job in the United States and woman is
1:16:57losing more than that that's why the
1:16:59total is really you know smaller than
1:17:01the loss in woman so when you think
1:17:04about that what's happening now this is
1:17:06a little bit different than the current
1:17:08Paradox because this is more a deep
1:17:12marketization story so it's like
1:17:15suddenly the market shter was being shut
1:17:18down school closed that's no choice all
1:17:23the children have to come home
1:17:25so women have to do this home production
1:17:28not because you see the whole curent
1:17:30cator a different thing is saying that
1:17:32when it's cheaper to do home production
1:17:35you know when there's washing machine
1:17:37when there you don't need as much time
1:17:40why do you spend more time that was the
1:17:42co Paradox because they increase the
1:17:44total demand the co is a different thing
1:17:47the co is that is really a reverse of
1:17:51what we have been studying here there is
1:17:53a demarkation
1:17:55that's you know you you you your service
1:17:58sector is being shut down so this flow
1:18:01here that I was showing earlier on this
1:18:04flow so this thing is just shut down
1:18:08there's no tfp growth in the market
1:18:10anymore this is completely shut down
1:18:12hours came back okay so that's one of
1:18:15the reason why we see this big loss in
1:18:19women's hour and it's not just in United
1:18:21States the io report show that around on
1:18:25the Grobe major loss in our world is on
1:18:28woman and young people so that is the
1:18:30First Source that's a Dem martiz is
1:18:33basically Force back the second reason
1:18:36is when these sector shut down exactly
1:18:39as you point out these are the sector
1:18:42that hire mostly women you know the
1:18:44child care sector the restaurant sector
1:18:46the hotel sector they are the sector
1:18:48that create the job for woman so when
1:18:51they shut down that basically destroy
1:18:54the job for woman so that second
1:18:57application we were talking about that's
1:18:59the sector that intensively hire woman
1:19:02so essentially the co you know
1:19:05marketization is creating two job for
1:19:08woman the co which is a Dem
1:19:11marketization is destroying two job for
1:19:13women woman who work in hotel lost their
1:19:16job women who don't work in hotel like
1:19:19you know for myself for example if the
1:19:22school continue to close like that I
1:19:24need to rethink whether I can still
1:19:26continue my job at LSG because I have
1:19:29this son constantly at home that I have
1:19:32to cook and look after okay so this is
1:19:35the second channel so the de Dem
1:19:38marketization the co it's like
1:19:40destroying the two job that the
1:19:42marketization created so it's a very
1:19:45very important thing and this is very
1:19:47timely that's why you see over the media
1:19:49they were saying all the progress we
1:19:52made in the last 20 30 years on the
1:19:56gender equality in labor market could be
1:19:59destroyed by covid and for me that
1:20:02really bite is this Dem marketization
1:20:16process Jo you muted yes yeah any more
1:20:20questions from the front
1:20:21row uh yes I have a question
1:20:26yeah okay yeah a really nice
1:20:28presentation thanks Rael um so I my
1:20:31question is that because one of my
1:20:33research we found the so the black Fe
1:20:36female they're just working in the labor
1:20:38market much more than Whit female
1:20:42especially after the child birth so I'm
1:20:44just thinking this become a POS for me
1:20:47because when you think about the white
1:20:49female they have a much higher
1:20:51opportunity cost because they have a
1:20:52much higher rage in the labor market
1:20:55and also in term of the family budget
1:20:57they are much richer for this
1:21:00Outsourcing and but then in the end the
1:21:02result is that they're like at home much
1:21:06more than the black female so is there
1:21:09any idea you think that conv yeah I I
1:21:12think when you look at this I would say
1:21:15it would be a lot due to the income
1:21:17effect of the household so you might
1:21:19want to look at the Sorting Channel okay
1:21:22like you know what it's exactly like you
1:21:24point out they have more resources to
1:21:27Outsource but on the other hand it might
1:21:29be they also have more resources that
1:21:31allow the white female to stay home to
1:21:36spend time on child care so the Reon
1:21:38work by Ry uh Ry and Ry so they
1:21:42basically find this rock race that they
1:21:44find that in the US time news you see a
1:21:48much it's a big increase in the time
1:21:51span on child care because of the
1:21:54difficulty of getting in college so that
1:21:56could be that there would be a
1:21:58reallocation in the sense that they will
1:22:00now can afford to have the mother
1:22:04specialize in child care in order to you
1:22:07know provide for because Child Care is
1:22:09quite a difficult thing to our source in
1:22:12a
1:22:12way you know so this is a very stylized
1:22:16model think of all home production as
1:22:19one but even in the time news table you
1:22:22see earlier on even when many other home
1:22:24home production outsourced you will
1:22:26always have this big chunk which is to
1:22:29do with car that is quite difficult so I
1:22:32would say that the income effect will
1:22:34play quite a role on that Dimension and
1:22:37also historically it's the same thing
1:22:40it's not just I guess your data is more
1:22:42recent right but if you look at the
1:22:44historical data in the 19 and early 20th
1:22:48century you see the same thing the
1:22:50female labor force participation is much
1:22:52higher for the black female than the
1:22:54white female so I I think there is
1:22:57something to do with that there could
1:22:59also you know something about sorting
1:23:01and about the household I think that
1:23:03also could be um something to do with
1:23:07social Norm that people have highlighted
1:23:09you know doing the you know um early U
1:23:13maybe 1960 1750 you know the perfect
1:23:18housewife among the whites Community is
1:23:21someone who stay home so I think so so
1:23:24nor differences the attitude about woman
1:23:28working there it could be different I
1:23:31think if you look at man you might find
1:23:34a mirror image I I don't know but if you
1:23:38look at it I would suspect when you look
1:23:40at black man versus white man you might
1:23:44find a reverse pattern there yeah and
1:23:47also the skill Dimension might have to
1:23:50do a lot to it
1:23:53too yeah great thank
1:24:03you any other questions from the front
1:24:06row I have a
1:24:08question um thank you richel for your
1:24:11presentation and you mentioned in your
1:24:13slide page 35 that in the early stage of
1:24:17development female employment is falling
1:24:21and you mentioned this is the same for
1:24:23China okay provide more intuitions
1:24:26behind that because based on my personal
1:24:29observation I I came from China but I
1:24:32don't I I didn't say that it's like
1:24:35confused for me thank you so in terms of
1:24:39data um I only look at the IO Data so I
1:24:43you know I haven't followed the female
1:24:46uh employment uh data in the China per
1:24:50se so my claim was just I was looking at
1:24:53the I o employment data and what I see
1:24:57is in many lowincome country female
1:25:00employment was falling as they develop
1:25:03so the number I see for China and India
1:25:06was since
1:25:081990 so there's been a fall in female
1:25:11employment in
1:25:121990 until now in China and India now
1:25:17the reason why H you know in China
1:25:21particularly were fall one the story we
1:25:24try to propose here is based on the idea
1:25:27of family
1:25:28farm so as you know China um in
1:25:341978 we have the household
1:25:36responsibility system okay so all the
1:25:40household they're responsible for plot
1:25:42of land okay so essentially when you
1:25:46think about employment of women they
1:25:48might not be getting a salary job okay
1:25:51so it's not the employment is not just a
1:25:54wage job not that they are hiring in the
1:25:56factory or whatever they can be hire in
1:26:00the Family Farm not getting a wage but
1:26:03they get the output of the farm so that
1:26:06is the international definition of
1:26:08employment it doesn't have to be wage
1:26:10payment okay it can be profit or output
1:26:14so family business is a form of
1:26:17employment so I ILO will call it
1:26:20contributing family worker okay so your
1:26:23husband own the farm or your husband own
1:26:27a workshop you help out you are employed
1:26:30okay even though you don't get wage so
1:26:33that's the same phenomenon in China
1:26:35because the agriculture employment share
1:26:38in China is still very high in the 1980
1:26:41is about 70% people are higher in the
1:26:44Agriculture and that is a similar
1:26:47pattern across the world during the
1:26:48early development stage more than 50% of
1:26:51the labor force are higher in
1:26:53agriculture and agriculture in early
1:26:55development stage very often consists of
1:26:59Family Farm okay and so what is
1:27:03happening is as family farm become less
1:27:07important think about the case of china
1:27:10um a lot of people they would have um
1:27:13sell off uh will have leave to the city
1:27:16to work um you know leave the farm you
1:27:19know for the older parents there so they
1:27:21don't work on the farm and if you think
1:27:24about the doing as the economy develop
1:27:27what happen is Farm become bigger okay
1:27:30so look at the biggest difference become
1:27:33between farming in the UK versus
1:27:35Charming versus China is that in the UK
1:27:39you will have these very porh people who
1:27:41might have a degree from Oxford you know
1:27:44who are farmer but what they do is they
1:27:47have big farm lots of machine uh other
1:27:50people working for them they they they
1:27:52it's not like the farming you think of
1:27:54in the Family Farm in the China setting
1:27:57or in an African country so as the
1:28:00modernization happen where family farm
1:28:03is taken over by bigger more efficient
1:28:07Farm then this woman are release from
1:28:11this family farm so what the immediate
1:28:15response is they might not immediately
1:28:18join the labor market to go to other
1:28:21sector like manufacturing sector or
1:28:23services then first you know they have
1:28:26to find a way to cope with the home
1:28:27production Family Farm allow them to
1:28:30cope with home production quite flexibly
1:28:32they can stay at home or even family
1:28:34business so that might be the first
1:28:36reason so they just stick with a
1:28:38full-time housewife instead of looking
1:28:40for job second it could be that during
1:28:43early development stage the market
1:28:46sector is not that well de develop so
1:28:49they couldn't just send their kid to a
1:28:52school or to a child care center
1:28:54okay so they do have to stay behind to
1:28:56look for that and also if the market
1:28:59service sector are not well developed
1:29:00they cannot find the job easily either
1:29:03okay so this could be the reason why
1:29:06during early development stage as family
1:29:09farm disappear we see female employment
1:29:13going down okay so this is one story we
1:29:15provide in this paper there could be
1:29:18other story you know some people have
1:29:20argu about in the case of United States
1:29:24for example they say oh the and there
1:29:27are social Norm against wom going into
1:29:30dirty Factory okay they don't uh the
1:29:34family don't want the wife to go into
1:29:37dirty Factory to work or go outside the
1:29:39home you know so the social Norm could
1:29:41play an important role to so as the
1:29:45demand for the family farm strength they
1:29:47couldn't go out to other sector of the
1:29:50economy so this could also be the reason
1:29:53and also Al people have been talking
1:29:55about income effect we kind of related
1:29:57to what I reply to ji so they could say
1:30:01that well because um as family get
1:30:04richer they are released from that
1:30:06subsistent need for agriculture when you
1:30:10are very poor you really need to work
1:30:12because you need to have some subsistent
1:30:14consumption but once you come out of
1:30:16that stage then the income effect kick
1:30:19in it allow you to work less okay and
1:30:22perhaps in enjoying Leisure and other
1:30:25things so these could all be factors how
1:30:27you generate the decline in the
1:30:29employment as economy develop Rachel yes
1:30:35I wanted to ask one more uh question
1:30:37that's from the Q&A if you could just
1:30:38briefly answer though because we're
1:30:40basically at a time okay the question is
1:30:43from stepen Barlo have you seen any
1:30:45particularly interesting or odd Trends
1:30:48well first he thanks you but then he
1:30:49says have you seen any particularly
1:30:50interesting or odd Trends if you look at
1:30:53disaggregating by class rather than
1:30:56gender class what class sorry like
1:31:00social class I don't know if that would
1:31:02be by income level or by um education
1:31:06level yeah so I guess my uh ML on the
1:31:10skill could be one you know dimension on
1:31:14the education level there a skill and
1:31:16also their occupation but uh when they
1:31:20say unusual Trend I I'm not sure I
1:31:23understand but there are a quite a lot
1:31:25of research looking at by skill group
1:31:28and there also like ji point out there's
1:31:30also Trend by race okay different race
1:31:34how they perform um and by occupation so
1:31:38that that's all I can
1:31:41say okay I'm not sure I answer the
1:31:43question very well but yeah um given the
1:31:47time we had that's probably ideal um I
1:31:50wanted to thank you Rachel um as I've
1:31:52mentioned to the
1:31:54each class the these professors are
1:31:56doing this out of the kindness of their
1:31:57heart um I think that the turnout we
1:32:00have shows the appreciation for your
1:32:02work Rachel so thank you very much I
1:32:05also wanted to correct um that I said
1:32:08next week but next week is actually the
1:32:10spring break of the course so you can
1:32:12all have fun vacationing on the beach uh
1:32:15and but the following week on Thursday
1:32:18and I guess that's March 11th and 12th I
1:32:21think uh we have Mark Rosen swag talking
1:32:24about Labor Market frictions in
1:32:25developing countries and Richard
1:32:27Rogerson talking about misallocation so
1:32:30we look forward to seeing all of you uh
1:32:32in two weeks thanks again Rachel thank
1:32:35you very much thank you everyone for
1:32:37coming I hope you enjoy
1:32:39bye thank you thank you thanks thank you
1:32:44thanks Jo you you thank you everyone bye
1:32:48thanks Rachel I hope I I my my memory of
1:32:52your paper was big I hope it was
1:32:54correctly represented I think so too but
1:32:57it was a long time ago yeah it is indeed
1:33:00right I was still a graduate student
1:33:02when you that home production paper and
1:33:04GDP accounting yeah no thank you though
1:33:08thanks for remembering it thank you
1:33:09everyone I hope you enjoy okay
1:33:22bye
1:33:52for for

About this transcript

This page contains the full transcript of Lecture 4: Structural transformation, home production, and labour markets by CEPR & VideoVox Economics, generated from the public captions YouTube serves with the video. The transcript has 13,843 words across 2,000 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.