Labor Market Development

Ping WangDepartment of Economics

Washington University in St. Louis

March 2017

1

A. Introduction

The labor market plays a key role in the process of economic development. Thisincludes many factors underlying the formation of human-related capital, oftenreferred to as "embodied" capital.

There are many dimensions along which labor market decisions and performanceinteract with the development process. We will discuss some selected topics, basedon the modern literature summarized below:

! Educational Choice: Lucas (1988), Laing-Palivos-Wang (1995), Fender-Wang (2003)

! The Role of Teachers: Tamura (2001), Chetty-Friedman-Saez-Turner-Yagan (2017)

! On-the-Job Learning: Lucas (1993), Laing-Palivos-Wang (2004)! Entrepreneurship: Bernhardt & Lloyd-Ellis (2000), Jiang-Wang-Wu

(2009)! Fertility and Labor Trade-off: Becker-Murphy-Tamura (1990), Scotese-

Wang-Yip (1994), Galor-Weil (2000),Greenwood-Seshadri (2005)! Health Capital: Acemoglu-Johnson (2007), Wang-Wang (2013) ! Human Capital Stratification: Benobou (1996), Chen-Peng-Wang (2009)! Inequality: Glomm-Ravikumar (1992), Acemoglu-Dell (2009)

2

B. The Role of Teachers: Tamura (2001)

! Empirical facts of Schooling across U.S. States: 1901-90" enrollment rate (73.3 to 92.1%): by 6% over 1901-60; 12% over 1960-90" class size (36.9 to 16.9 students/teacher): by 12 1901-60 & 8 over 1960-90" relative teacher salary (from 1.53 to 2.35 to 1.76 teacher to average income

ratio): by 0.8 over 1901-60 and 0.6 over 1960-90

1. The Model

! Two-period lived overlapping generations with constant population

! Altruistic Preferences: U = , 0< β < 1 and σ < 1

! School Quality and Human Capital Evolution:" teacher quality (teacher-parents human capital ratio):

" class size (student-teacher ratio):

! Human capital accumulation (HC): ,

3

! Individual Budget Constraints (BC): ! Local Governments' Budget Constraints (GBC):

" poor school districts (NPt = α):

" rich school districts (NRt = 1-α):

2. Equilibrium

! Theoretical results: under ε < 1, human capital is rising over time, with thepoor districts growing at faster speed than rich districts (convergence)

! Empirical findings: " C reduces real per capita income growth, while Q enhances it" Over the entire sample (1882-1990),

- enhancement in Q accounts for 60% of real growth- reduction in C accounts for 40%

! In the past 4 decades (1950-1990), " enhancement in Q accounts for 13% of real growth " reduction in C accounts for 85%

4

3. College Admission and Intergenerational Mobility: Chetty-Friedman-Saez-Turner-Yagan (2017)

! Administrative data (U.S., 1999-2013) with over 30 million college students! Large variation in parental income (chart from NY Times)

5

! Top 38 private universities/colleges had more students from top 1% ($630K)parental income than bottom 60% (<$65K)

% students fromSchools top 1% bottom 60%

1. Washington University in St. Louis 21.7 6.12. Colorado College 24.2 10.53. Washington and Lee University 19.1 8.44. Colby College 20.4 11.15. Trinity College (Conn.) 26.2 14.36. Bucknell University 20.4 12.27. Colgate University 22.6 13.68. Kenyon College 19.8 12.29. Middlebury College 22.8 14.210. Tufts University 18.6 11.814. Vanderbilt 22.8 14.916. Dartmouth College 20.7 14.421. Princeton University 17.0 13.626. Duke University 19.2 16.527. Yale University 18.7 16.329. University of Pennsylvania 18.7 16.530. Brown University 19.5 18.2

6

! Conditional on college attended, children from low and high income familieshave similar earning outcomes

! Mobility rates (from bottom 20% to top 20% or top 1%) varies across schools:mobility rates = college access rates x success rate" access rate = fraction of students from bottom distribution (e.g. quintile)" success rate = fraction of students reaching top distribution at work

7

! Test performance (SAT), private school selectivity (reject rates), college costs,STEM shares and faculty quality (salary) all reduce the access rate but raisethe success rate, leading to greater intergenerational mobility (especially forupper tail mobility)

! Again, the positive role of teachers is confirmed at the college level! Open issues:

" private vs. public education" need vs. merit based scholarship

8

C. On-the-Job Learning

! Learning on-the-job can enhance productivity significantly, at least over ashorter horizon

! Using the Liberty ship (same blue print) data, Searle (1945) and Rapping(1965) identified 12-24% and 11-29% learning-by-doing effect in production

! In reality, product cycles are shortened over time, making pure learning on-the-job less productive

1. The Model: Lucas (1993)

! Final good output:

! Experience accumulation: " experiences grow over time" more work time helps accumulating experiences" while A is a conventional technical factor, the overall productivity depends

on the experience input

9

2. Equilibrium

! Set =>

! Productivity growth: if

3. Main Results

! Presence of scale effect:

! Based on Rapping (1965), ξ = 0.2; so, " learning on-the-job faces rapid decay" sustained human-capital growth requires:

- education- retraining- beyond-the-job learning

10

D. Entrepreneurship

! Cagetti and De Nardi (2006): based on the Survey of Consumer Finances,entrepreneurs measured by self-employed business owners account for only7.6% of the U.S. population but for almost 1/3 of the total net worth

! Mondragon-Velez (2006): entrepreneurs receive more than 20% of income ofthe entire population

! Entrepreneurship:" The International Social Survey Programme of 1989 shows about 63% of

Americans, 48% of Britons and 49% of Germans desire to becomeentrepreneurs

" Only about 15% realize their dream

11

! Major factors affecting agent's choice (Kihlstrom-Laffont (1979): " entrepreneurial ability vs. labor skills" access to capital markets" individual attitude towards risk

! Barriers to entrepreneurship: " preference bias toward "heroes" limited by true ability

(Blanchflower-Oswald 1998)" financial/liquidity constraints (Evans-Jovanovic 1989, Evans-Leighton

1989, Den Haan-Ramey-Watson 2003)! Theoretical frameworks:

" Lucas (1978), Bernhardt & Lloyd-Ellis (2000): model entrepreneurs asmanagers

" Jiang-Wang-Wu (2009a,b): model the 4 special features of entrepreneurs,entrepreneurial ability, financing, risk-taking and heroic preferences- presence of selectivity, scale and loanable fund supply effects- tightened credit market =>

# more selective high-quality entrepreneurs (+)# less entrepreneurial scale (-)# more loanable fund supply (+)

- entrepreneurship and growth need not be positive related

12

E. Fertility and Labor Trade-off

! Long-term trend in population (Western Europe, Maddison 1982/1995):" 500-1500: rising pop, no per capita output growth" 1500-1870: rising pop, rising per capita output" after 1870: declining pop, rising per capita output

! Fertility decline since mid-1800:" reducing infant mortality enables lower fertility given the same desired

quantity of children" rising income makes education more affordable and encourages the trade-

off of quantity for quality in childbearing" rising opportunity cost increases childbearing cost

! Roles of demographic transition played in economic development: high fertilityis associated with low development (Malthusian trap)

! The modern literature includes:" quality-quantity trade-off: Razin&Ben-Zion (1975), Becker-Barro (1989),

Scotese-Wang-Yip (1994)" fertility and human capital: Becker-Murphy-Tamura (1990):" Galor-Weil (2000): demographic transition from Malthusian to modern" Greenwood-Seshadri (2005): baby boom/baby burst

13

! The long fertility cycle in the U.S.:

14

1. The Model: Scotese-Wang-Yip (1994)

! Households has endogenous consumption-saving, labor-leisure andchildbearing/childrearing decisions, where childbearing/childrearing requirestime and resources cost

! The time constraint (TC): ! The budget constraint (BC):

! Optimization: max s.t. (TC) and (BC)

2. Equilibrium and Results

! Fertility declines in response to a preference shift away from the quantity ofchildren (μ), higher time costs in fertility (s) and higher labor productivity (A)

! Empirical findings:" in response to a preference shift away from fertility, fertility declines and

output rises" preference-driven fertility shocks explain about 80% of movements in

fertility and 25% of movements in output

15

F. Health Capital

! Sub-Saharan Africa and South Asia have suffered high disease and intensepoverty. Poor health environments may be important for explaining why"geography" matters for growth, especially for those countries in sub-SaharanAfrica and South Asia long falling in the low-growth trap.

! Basic idea:" Increased life expectancy raises

population and lowers capital-labor and land-labor ratios,leading to lower per capita output

" Lengthened life expectancyencourages labor-marketparticipation and saving,resulting in more capitalaccumulation and higher percapita output

" This "non-monotone" effects canbe best seen from experiencesfacing initially poor countries

16

1. The Model: Acemoglu-Johnson (2007)

! Country i’s aggregate output: ! Land: ! Effective labor: ! Life expectancy Xit, affecting:

" Population and technology: and " Individual human capital:

! Capital accumulation with an exogenous saving rate s: + (1 - δ)

2. The Estimation

! Regression:

depending on life expectancy and an array of other variables

17

3. Data

Life Expectancy Initially Poor

Initially Middle-Income

InitiallyRich

At Birth in 1900 28.77 36.92 49.36

At Birth in 1940 40.63 50.93 65.13

At Birth in 1980 61.92 69.66 74.3

At Age 20 in 1940 56.96 64.51 70.41

At Age 20 in 1980 70.27 73.59 75.73

4. Main Findings

! Predicted mortality has a large effect on changes in life expectancy since 1940,but not before

! 1% increase in life expectancy raises population by 1.7-2%! The effect of life expectancy on per capita real GDP is negligible

18

5. Health and Development Accounting: Wang-Wang (2013)

! An organizing framework:" Production: , where

" Human capital: where h measures physical health and m measures mental knowledge

" Mincerian equation: " log calculus => growth accounting:

where, in addition to the residual TFP (the constant term) output growth can be decomposed into 3 components:

- capital accumulation- education enhancement- health improvement

19

! Data:

20

21

22

23

24

25

26

27

! Results: chance to pull out of poverty trap (41 currently trapped economies)

28

G. Aged Society

! While most developedcountries and manydeveloping countries havecompleted demographictransition, their newdemographic challenge israpid ageing as a result oftechnological change inhealth and medicine andrising income, among others.

! Declined fertility andlongevity imply highermedian age (based on UN2015 data/projection)

! The dependency ratio rises:the ratio in EU15 is 1/1.8 in2015, projected to rise to1/1.2 in 2065 (World Bank)

29

! World population by age group (Bloom-Canning-Flink 2011, based on UN)

30

! Share of 60+ population in the world by regions (ibid)

31

! Share of 60+ (blue) and 80+ population in 4 large developing economies (ibid)

32

! City labor participation in the largest economy, China (ibid)

33

! Issues:" lower labor force participation" lower productivity on per

capita basis" higher health care and long

term care cost" more welfare spending" less pension sustainability" lower growth: World Bank

projection – lower by 0.4% inthe next 50 years)

! Policy:" raise retirement age" encourage adequate increase in fertility" enhance immigration" promote healthy ageing" improve pension management: from pay-as-you-go to fully-funded or to

even more demographic-considered flexible system and to secure decentreturn

34

H. Locational Human-Capital Mobility: Bond, Riezman and Wang (2010)

! Along the global trend of economic development, it is often observed rapidindustrial transformation accompanied by continual rural-urban migration(Lewis 1954, Fei-Ranis 1961, 1964, and Sen 1966)

! In many developing countries there are yet abundant supplies of surplus labor:" urbanization trend (% rural population):

" Key features:- abundant rural surplus labor (13-28% of total population)- large urban-rural unskilled wage gap (2:1)- low growth in real unskilled wage (less than 1% by excluding labor-

augmenting technical progress)- gradual upgrade of some unskilled migrants to skilled (Lucas 2004)

! Conventional literature is either static (Todaro 1969, Hariss-Todaro 1970) ordynamic without modeling explicitly the dynamic process of labor migration(Drazen-Eckstein 1988, Glomm 1992, Bencivenga-Smith 1997, Banerjee-Newman 1998, with Lucas 2004 as an exception)

35

! The recent rise of the emerging economies in Asia and Latin America has beenaccompanied by rural-urban migration and trade liberalization

! Question: what is the role of open-trade policy played in facilitating locationalhuman capital mobility and long-run growth

! Responses to tightened trade protection (higher import tariff) when the importsubstitution sector Y is more capital-intensive than the exporting sector:" protection raises the capital-labor ratio in Y (kY) more than that in X (kX)" output in Y rises whereas output in X reduces; aggregate output decreases" both capital and urban employment are lower (urbanization )" economic growth is lower

! Similar results in response to barriers to rural-urban migration, where morebarriers reduces urbanization and growth" policies raising such barriers:

- China: household registration system- India: cast system

" policies reducing such barriers:- Taiwan: public training programs & job search assistance- UK (London)/US (Chicago)/China (tier-1 cities): public

provided/subsidized urban housing

36

! An organizing framework: rural to urban migration continues until theexpected rural and urban wages (or utility) are equalized

! More barriers or more costly to move to urban areas will lead to a higherability cutoff, resulting in slower urbanization

37

! Transition to urban society: Regimes 1 (1980-94), 2 (1995-2001), 3 (2002-08)

" migration cost reduction and TFP most important for earlier growth;skill/service expansion and pop control most crucial for later

" urbanization driven by tariff/migration cost reduction and TFP earlier, bypop control and service expansion later

38

0 . 3 9 o r l o w e rD a l l a s , S e a t t l e , P o r t l a n d

M i a m i , D e n v e r , S a c r a m e n t o , O r l a n d o

S a n D i e g o , N o r f o l k , S a n F r a n c i s c o 0 . 4 0 - 0 .4 9

T a m p a , S a n A n t o n i o , P h o e n i x , M i n n e a p o l i s

H o u s t o n , B o s t o n , L o s A n g e l e s0 . 5 0 - 0 .5 9

P i t t s b u r g h , A t l a n t a , K a n s a s C i t y

P h i l a d e l p h i a , C i n c i n n a t i , C h i c a g o , I n d i a n a p o l i s0 . 6 0 - 0 .6 9

M i l w a u k e e , C l e v e l a n d , S t . L o u i s , N e w Y o r k

0 . 7 0 o r h i g h e rD C - B a l t i m o r e , D e t r o i t

D i s s i m i l a r i t y I n d e xM e t r o p o l i t a n S t a t i s t i c a l A r e a ( M S A )

0 . 3 9 o r l o w e rD a l l a s , S e a t t l e , P o r t l a n d

M i a m i , D e n v e r , S a c r a m e n t o , O r l a n d o

S a n D i e g o , N o r f o l k , S a n F r a n c i s c o 0 . 4 0 - 0 .4 9

T a m p a , S a n A n t o n i o , P h o e n i x , M i n n e a p o l i s

H o u s t o n , B o s t o n , L o s A n g e l e s0 . 5 0 - 0 .5 9

P i t t s b u r g h , A t l a n t a , K a n s a s C i t y

P h i l a d e l p h i a , C i n c i n n a t i , C h i c a g o , I n d i a n a p o l i s0 . 6 0 - 0 .6 9

M i l w a u k e e , C l e v e l a n d , S t . L o u i s , N e w Y o r k

0 . 7 0 o r h i g h e rD C - B a l t i m o r e , D e t r o i t

D i s s i m i l a r i t y I n d e xM e t r o p o l i t a n S t a t i s t i c a l A r e a ( M S A )

I. Human Capital Stratification

! In reality, households are stratified in various degrees by race, income,education and other socioeconomic indicators

! The Dissimilarity index (Duncan-Duncan 1955): using the 2000 Census data,most of the 30 largest Metropolitan Statistical Areas were highly stratified:

39

! It has been shown that since 1980, racial segregation in the U.S. has declinedwhile economic segregation has risen.

! Human capital and housing are believed the two primary sources of economicsegregation.

1. The Model: Benobou (1996)

! Interactions " Local positive spillovers - in human capital evolution" Global positive spillovers - in goods production (as in Lucas 1988)

! Human Capital and Education" human capital evolution: h u h Et

i iti

ti

ti

1

11 (( ) ) ( )" public education: E y dG yt

iti

ti

ti

ti ( )

! Output: y H hti

t ti

1

1( ) ( )

! Combining the above relationships , h h H Lt

i iti

t ti

11 1 1 ( ) ( ) ( )( ) ( )( )

where is a "local" human capital aggregator Li

40

2. Segregated vs. Integrated Equilibrium

! Segregated equilibrium features locational clustering by human capital/income! Integrated equilibrium features mixture of groups with different human

capital/income! Two fundamental forces:

" complementarity between and hi => segregation (assortativeLimatching)

" complementarity between and hi => integration (homogenizing)H

3. Results

! Co-existence of segregated and integrated equilibria! Integration lowers inequality as compared to segregation! Integration lowers growth in SR but raises it in LR, because H has a larger

scale effect in the long run

41

J. Informality

! The presence of an informal sector is a world-wide phenomenon particularly indeveloping countries with larger tax burden, heavier labor-market barriersand restrictions and lower government institution quality (de Soto 1989):" high business taxes" labor regulations (work hour restrictions, overtime pay requirement,

minimum wage, fringe requirements such as health insurance, sick leave,maternal leave and pension, severance pay)

" consumer protection" environmental regulations" high access costs to legality and entry, including financial cost and red tape

! While the informal sector can be a tax avoidance heaven (Loaya 1996), it mayalso serve as a temporalizing state to absorb urban unemployment (Marjit-Wang 2014)

! Informality also incurs costs:" penalties including fines and forced closure" bribes to corrupt government officials (10-15% of gross income, de Soto

1989)" inability to benefit from government-provided services" inability to access to formal capital markets

42

! The size of the informal sector (Loayza 1996)

43

! Informality and development (ibid):

" informality has direct and indirect (via tax-financed public infrastructure)effects on economic growth and development

44

! Employee characteristics between formal the informal sectors: a case study ofBuenos Aires, 1993-95 (Amaral-Quintin 2006)

" informal sector employees are younger and less educated, with shortertenure

" they earn 25% less" they are in much smaller firms

45

! Organizational choice based on establishment size s, revenue R, production cost

C, other legal/regulatory cost X, fine φ, bribe β, corporate income tax rate τ

and public infrastructure benefit g:

" informal firm profit: Π(s; 0) = (1-β-φ)[R(s) - C(s)]

" formal firm profit: Π(s; 1) = (1-τ)[R(s) - (1-g)C(s) - X]

" larger firms are more willing to be legalized (fixed cost X less important)

" higher tax, lower infrastructure benefit and more legal/regulatory costs all

encourage firms to become informal

" costly bribes and higher fines discourage informality

46

K. Income Inequality Across Space and Time

! Stylized fact: large cross-country and within-country differences in per capitaincome

! Potential causes of such disparities:" differences in human capital" differences in technological know-how" differences in production efficiency due to various institutions and

organizations

1. The Model: Acemoglu-Dell (2009)

! Measure of inequality (municipal m in country j) by the Theil index:

where is the within-municipal m Theil index in country j! Alternative measures: mean log deviation, variance/coefficient of variation, gini

coefficient

47

! Wage inequality

90/10Theil index

BetweenCountry

WithinCountry

Municipals

actual pop weights 34.2 0.25 0.544 equal pop weights 28.6 0.285 0.622Regions

actual pop weights 36.7 0.203 0.529

equal pop weights 32.7 0.139 0.615

" more within than between country inequalities" more inequality using municipal than region data

48

! Decomposition of wage inequality measured by Theil index

Overall Inequality Residual Inequality

BetweenCountry

BetweenMunic.

WithinMunic.

BetweenCountry

BetweenMunic.

WithinMunic.

Municipals

actual pop weights 0.265 0.067 0.424 0.033 0.04 0.389 equal pop weights 0.301 0.105 0.474 0.041 0.053 0.404 U.S. 0.05 0.365 0.02 0.291

" "residual" within-the-skilled-group inequalities account for a large portionof overall inequalities

" within-municipal disparities are most important for wage inequalities" between-country disparities are important only for "non-residual"

between-skilled-and-unskilled-group inequalities" between-municipal disparities are never important

49

L. The Battle between the Top 1% and the Remaining 99%

! Income inequality

! Wealth inequality" U.S. Wealth Inequality: https://www.youtube.com/watch?v=QPKKQnijnsM

! Capital In The 21st Century: " BBC: https://www.youtube.com/watch?v=HL-YUTFqtuI" ABC: https://www.youtube.com/watch?v=I05wLUuvQGM

50

M. Human Capital Policy in Developing Countries

! Mandatory education: 6, 9 or 12 years of publicly provided education! Vocational schools, particularly tech schools and commercial colleges! Case studies:

" Brazil (Narayan, Patel, Schafft, Rademacher, and Koch-Schulte 2000): lowteacher salaries, insufficient schooling equipments/supplies, andcorruption have wrecked incentive for quality education, with- teachers selling publicly provided textbooks and supplies- school officers receiving bribes from building construction and

equipment purchases- both teachers and staff earning from secondary jobs- low school attendance without monitoring

" Pakistan (Husain 1999): politicians dispense teaching positions aspatronage, leading to:- large fractions of unqualified teachers with 75% failing to pass the

exams taken by students- large-scale cheating at examinations supervised by those teachers- campus gangs, with high school students from rival religious fractions

fighting on campus using AK-47s (more guns than textbooks)

51

" Taiwan (Tallman-Wang 1994): thorough education reform wereundertaken in the 1950s and 1960s, with- 1-12 public school teachers being required to pass nationally

monitored qualification exams- 1-12 public school teachers being provided with good salaries,

superior fringes (health and pension) and job security- county-level standardized tests for students to take for entry to junior

high and senior high, and national university entrance exam- 1-9 mandatory education- strong tech, agricultural and commercial schools that were tied to

related businesses- publicly provided job advertisement and training