Motivation and Overview Two Stylized Facts A Model of Intergenerational Cultural Transmission

“Just Do Your Job”: Obedience, Routine Tasks, and the Pattern of Specialization

Filipe R. Campante Harvard Kennedy School & NBER

Davin Chor NUS

3 May 2018 National Taiwan University

Campante, Chor

Obedience, Routine Tasks, and the Pattern of Specialization

1 / 37

Motivation and Overview Two Stylized Facts A Model of Intergenerational Cultural Transmission

Motivation: Why “Obedience”? Obedience: Propensity to comply with orders from a source of authority. I

As a behavioral trait: Well-studied in psychology

I

A literature on “conformity” in sociology

Why might “obedience” be consequential for economic outcomes?

Campante, Chor

Obedience, Routine Tasks, and the Pattern of Specialization

2 / 37

Motivation and Overview Two Stylized Facts A Model of Intergenerational Cultural Transmission

Motivation: Why “Obedience”? Obedience: Propensity to comply with orders from a source of authority. I

As a behavioral trait: Well-studied in psychology

I

A literature on “conformity” in sociology

Why might “obedience” be consequential for economic outcomes?

Campante, Chor

Obedience, Routine Tasks, and the Pattern of Specialization

2 / 37

Motivation and Overview Two Stylized Facts A Model of Intergenerational Cultural Transmission

Motivation: Why “Obedience”? I

Ngai and Chan (2012) describe Foxconn’s production process as: “not [requiring] ‘skill’ or thought; only strict implementation of instructions from management and mechanical repetition of each simple movement are required.” (p.395) Workers are thus managed “through the principle of ‘obedience, obedience, and absolute obedience’.” (p.398)

I

One narrative in the context of East Asia: “Asian” or “Confucian” values were critical for mobilizing the workforce to achieve economic growth. “Harmony and cooperation were preferred over disagreement and competition.” (Huntington 1991, p.24)

Campante, Chor

Obedience, Routine Tasks, and the Pattern of Specialization

2 / 37

Motivation and Overview Two Stylized Facts A Model of Intergenerational Cultural Transmission

This project: Overview I

Study one specific aspect of cultural attitudes relevant to workplace productivity: I

“Obedience in the Workplace” (from the World Values Survey)

I

Propensity to follow instructions vs question them in a work environment

Campante, Chor

Obedience, Routine Tasks, and the Pattern of Specialization

3 / 37

Motivation and Overview Two Stylized Facts A Model of Intergenerational Cultural Transmission

This project: Overview I

I

Study one specific aspect of cultural attitudes relevant to workplace productivity: I

“Obedience in the Workplace” (from the World Values Survey)

I

Propensity to follow instructions vs question them in a work environment

Observation #1: Cultural attitudes that promote worker productivity in some activities (e.g., assembly line manufacturing) may not be conducive Quote in others (e.g., computer coding, research)

Campante, Chor

Obedience, Routine Tasks, and the Pattern of Specialization

3 / 37

Motivation and Overview Two Stylized Facts A Model of Intergenerational Cultural Transmission

This project: Overview I

Study one specific aspect of cultural attitudes relevant to workplace productivity: I

“Obedience in the Workplace” (from the World Values Survey)

I

Propensity to follow instructions vs question them in a work environment

I

Observation #1: Cultural attitudes that promote worker productivity in some activities (e.g., assembly line manufacturing) may not be conducive Quote in others (e.g., computer coding, research)

I

Observation #2: Culture is not immutable, but evolves endogenously, partly in response to economic conditions (Bisin and Verdier 2011) Growing body of evidence that features of the economic environment affect what cultural attitudes persist and get transmitted across generations (E.g.: Alesina, Giuliano, and Nunn 2013, Giuliano and Nunn 2016)

Campante, Chor

Obedience, Routine Tasks, and the Pattern of Specialization

3 / 37

Motivation and Overview Two Stylized Facts A Model of Intergenerational Cultural Transmission

This project: Overview Three components to this study: 1. Establish a “Specialization Fact”: How does Workplace Obedience relate to the pattern of specialization? 2. Establish an “Obedience Fact”: How does the (lagged) pattern of specialization in turn shape Workplace Obedience? 3. Develop a model to understand: How do attitudes towards obedience and the structure of the economy co-evolve?

Campante, Chor

Obedience, Routine Tasks, and the Pattern of Specialization

4 / 37

Motivation and Overview Two Stylized Facts A Model of Intergenerational Cultural Transmission

Overview: 1. “Specialization Fact” I

Hypothesize that: Pro-obedience workplace attitudes are beneficial to productivity in relatively routine tasks I

Turn to Autor, Levy and Murnane (2003)’s indices of task routineness, coded up from the Dictionary of Occupational Titles

I

Use export data at the country-industry level to capture specialization patterns

Campante, Chor

Obedience, Routine Tasks, and the Pattern of Specialization

5 / 37

Motivation and Overview Two Stylized Facts A Model of Intergenerational Cultural Transmission

Overview: 1. “Specialization Fact” I

I

Hypothesize that: Pro-obedience workplace attitudes are beneficial to productivity in relatively routine tasks I

Turn to Autor, Levy and Murnane (2003)’s indices of task routineness, coded up from the Dictionary of Occupational Titles

I

Use export data at the country-industry level to capture specialization patterns

“Specialization Fact”: I

In countries where the workforce becomes (say) more pro-obedience, this is associated with a relative rise in exports for industries with a higher routine task content. As stated, this is a within-country finding.

I

Robust to controlling for the role of other cultural attitudes with which workplace obedience might be correlated (e.g., “hard work”, “work as a duty”, “individualism”)

Campante, Chor

Obedience, Routine Tasks, and the Pattern of Specialization

5 / 37

Motivation and Overview Two Stylized Facts A Model of Intergenerational Cultural Transmission

Overview: 2. “Obedience Fact” I

Hypothesize that: If the economy is oriented towards routine industries at time t, this incentivizes the adoption and transmission of pro-obedience workplace attitudes to generation t + 1 I

Use the micro WVS data.

I

Construct export-routineness (expRT ) to summarize how oriented the economy was towards routine tasks

Campante, Chor

Obedience, Routine Tasks, and the Pattern of Specialization

6 / 37

Motivation and Overview Two Stylized Facts A Model of Intergenerational Cultural Transmission

Overview: 2. “Obedience Fact” I

I

Hypothesize that: If the economy is oriented towards routine industries at time t, this incentivizes the adoption and transmission of pro-obedience workplace attitudes to generation t + 1 I

Use the micro WVS data.

I

Construct export-routineness (expRT ) to summarize how oriented the economy was towards routine tasks

“Obedience Fact”: I

More educated individuals less likely to agree that workplace obedience is important: ∂ObedWork < 0, . . . ∂Educ

I

But: This effect of education is dampened when the individual’s birth cohort is exposed to a greater degree of export-routineness during their ∂ 2 ObedWork formative years: ∂Educ > 0. ∂expRT Findings hold under an IV strategy, with country c’s export profile predicted using a “shift-share” variable based on world trade conditions Campante, Chor

Obedience, Routine Tasks, and the Pattern of Specialization

6 / 37

Motivation and Overview Two Stylized Facts A Model of Intergenerational Cultural Transmission

Overview: 3. Model These two Facts motivate an overlapping generations model where: I

The economy is comprised of a routine and a nonroutine sector; and

I

Parents decide how much Human Capital and Obedience to invest in their children (c.f., Bisin and Verdier (2001); Tabellini (2008))

Campante, Chor

Obedience, Routine Tasks, and the Pattern of Specialization

7 / 37

Motivation and Overview Two Stylized Facts A Model of Intergenerational Cultural Transmission

Overview: 3. Model These two Facts motivate an overlapping generations model where: I

The economy is comprised of a routine and a nonroutine sector; and

I

Parents decide how much Human Capital and Obedience to invest in their children (c.f., Bisin and Verdier (2001); Tabellini (2008))

Model delivers: I

Predictions on the determinants of transmitted obedience that are consistent with the “Obedience Fact”

Campante, Chor

Obedience, Routine Tasks, and the Pattern of Specialization

7 / 37

Motivation and Overview Two Stylized Facts A Model of Intergenerational Cultural Transmission

Overview: 3. Model These two Facts motivate an overlapping generations model where: I

The economy is comprised of a routine and a nonroutine sector; and

I

Parents decide how much Human Capital and Obedience to invest in their children (c.f., Bisin and Verdier (2001); Tabellini (2008))

Model delivers: I

Predictions on the determinants of transmitted obedience that are consistent with the “Obedience Fact”

I

In the long-run, the possibility of an “Obedience Trap”: Specializing in routine sectors entrenches a culture of following instructions, at the expense of expanding into nonroutine activities.

Campante, Chor

Obedience, Routine Tasks, and the Pattern of Specialization

7 / 37

Motivation and Overview Two Stylized Facts A Model of Intergenerational Cultural Transmission

Overview: 3. Model These two Facts motivate an overlapping generations model where: I

The economy is comprised of a routine and a nonroutine sector; and

I

Parents decide how much Human Capital and Obedience to invest in their children (c.f., Bisin and Verdier (2001); Tabellini (2008))

Implications: I

A cultural dimension to the phenomenon of labor market polarization? (e.g.: Autor and Dorn 2013, Goos and Manning 2007, Goos, Manning and Salomons 2014)

I

Cultural traits that serve a country well at early stages of development could ultimately hinder it from transitioning into more complex production Quote activities

Campante, Chor

Obedience, Routine Tasks, and the Pattern of Specialization

7 / 37

Motivation and Overview Two Stylized Facts A Model of Intergenerational Cultural Transmission

Roadmap for this talk

1. Motivation and Introduction 2. Two Stylized Facts 2.1 Data 2.2 The Specialization Fact 2.3 The Obedience Fact

3. Model: Endogenous cultural transmission 4. Conclusion

Campante, Chor

Obedience, Routine Tasks, and the Pattern of Specialization

8 / 37

Motivation and Overview Two Stylized Facts A Model of Intergenerational Cultural Transmission

Background on the Data Variables The “Specialization Fact” The “Obedience Fact”

Background on the Data

Campante, Chor

Obedience, Routine Tasks, and the Pattern of Specialization

9 / 37

Motivation and Overview Two Stylized Facts A Model of Intergenerational Cultural Transmission

Background on the Data Variables The “Specialization Fact” The “Obedience Fact”

The World Values Survey (WVS) I

Six waves (1981-2014)

I

> 300, 000 observations from 229 surveys (97 countries/territories)

I

Typical survey-wave: > 1000 respondents aged 15 and above

Question C061: [Obedience in the Workplace] “People have different ideas about following instructions at work. Some say that one should follow one’s superior’s instructions even when one does not fully agree with them. Others say that one should follow one’s superior’s instructions only when one is convinced that they are right. With which of these two opinions do you agree?” I

‘3’= Follow instructions; ‘2’= Depends; ‘1’= Must be convinced first

I

Available in Waves 1-5.

Countries

Campante, Chor

Obedience, Routine Tasks, and the Pattern of Specialization

10 / 37

Motivation and Overview Two Stylized Facts A Model of Intergenerational Cultural Transmission

Background on the Data Variables The “Specialization Fact” The “Obedience Fact”

Aggregating to the Country Level Needed: A country-level measure of workplace obedience attitudes at various points in time, t = 1985, 1990, . . . I

As a start: Can take a simple average of obedience scores for respondents from a country-cohort-gender bin, and then take a weighted-average of these based on the age structure of the workforce at time t

I

However: Prevailing economic conditions can affect reported responses E.g.: Someone currently working in an assembly line more likely to agree with following instructions

Campante, Chor

Obedience, Routine Tasks, and the Pattern of Specialization

11 / 37

Motivation and Overview Two Stylized Facts A Model of Intergenerational Cultural Transmission

Background on the Data Variables The “Specialization Fact” The “Obedience Fact”

Aggregating to the Country Level Needed: A country-level measure of workplace obedience attitudes at various points in time, t = 1985, 1990, . . . I

As a start: Can take a simple average of obedience scores for respondents from a country-cohort-gender bin, and then take a weighted-average of these based on the age structure of the workforce at time t

I

However: Prevailing economic conditions can affect reported responses E.g.: Someone currently working in an assembly line more likely to agree with following instructions

I

To address this: Extract a country-cohort-gender specific component ˆ cbg ) of reported workplace obedience attitudes. . . (D . . . after controlling for respondent observables (including employment status and occupation) and country-survey wave fixed effects

Campante, Chor

Obedience, Routine Tasks, and the Pattern of Specialization

11 / 37

Motivation and Overview Two Stylized Facts A Model of Intergenerational Cultural Transmission

Background on the Data Variables The “Specialization Fact” The “Obedience Fact”

Aggregating to the Country Level Estimate: ObedWorkr ,cbw = β0 + β1 Educr ,cbw + βX Xr ,cbw + Dcbg + Dcw + r ,cbw .

(1) Details

I

r : respondent; c: country; w : WVS wave

I

b: birth cohort (e.g.: 1965-1969, 1970-1974, 1975-1979,. . .)

I

Educr ,cbw : Education (From 1=‘Inadequately completed elementary education’; to 8=‘University with degree/Higher education - upper-level tertiary certificate’)

I

Xr ,cbw : Other respondent controls (number of children, marital status, employment status, occupation, size of town)

I

Dcbg : country-cohort-gender fixed effect

I

Dcw : country-survey wave fixed effect

I

r ,cbw : idiosyncratic noise Campante, Chor

Obedience, Routine Tasks, and the Pattern of Specialization

12 / 37

Motivation and Overview Two Stylized Facts A Model of Intergenerational Cultural Transmission

Background on the Data Variables The “Specialization Fact” The “Obedience Fact”

Aggregating to the Country Level Estimate: ObedWorkr ,cbw = β0 + β1 Educr ,cbw + βX Xr ,cbw + Dcbg + Dcw + r ,cbw .

(1) Details

I

Then compute: AvgObedWorkct =

X

ˆ cbg . ωcbgt D

(2)

(c,b,g )

where ωcbgt is the share of (c, b, g ) in the workforce aged 25-64 in year t. Figure

I

Caveat: Focus on within-country variation; view this as more meaningful than cross-country comparisons.

I

Verified that average responses are highly correlated within birth cohort, even if Figure surveyed at different ages.

Campante, Chor

Obedience, Routine Tasks, and the Pattern of Specialization

12 / 37

Motivation and Overview Two Stylized Facts A Model of Intergenerational Cultural Transmission

Background on the Data Variables The “Specialization Fact” The “Obedience Fact”

Task Routineness Premise: Obedience a complementary attribute for performing routine tasks. From Autor, Levy and Murnane (2003):

Details

I

Index measures (0-10 scale) coded up from the 1977 Dictionary of Occupational Titles (DOT), US Dept of Labor

I

Mapped to industry measures using weights from 1960 US Census PUMS

I

Five raw measures: I

T r ,c , routine cognitive: Set Limits, Tolerances, Standards

I

T nr ,c1 , nonroutine cognitive (interactive): Direction, Control, Planning

I

T nr ,c2 , nonroutine cognitive (analytic): Math

I

T r ,m , routine manual: Finger Dexterity

I

T nr ,m , nonroutine manual: Eye-Hand-Foot Coordination

Campante, Chor

Obedience, Routine Tasks, and the Pattern of Specialization

13 / 37

Motivation and Overview Two Stylized Facts A Model of Intergenerational Cultural Transmission

Background on the Data Variables The “Specialization Fact” The “Obedience Fact”

Task Routineness Premise: Obedience a complementary attribute for performing routine tasks. From Autor, Levy and Murnane (2003):

Details

I

Index measures (0-10 scale) coded up from the 1977 Dictionary of Occupational Titles (DOT), US Dept of Labor

I

Mapped to industry measures using weights from 1960 US Census PUMS

I

Compute three summary measures of industry task routineness, similar to Autor and Dorn (2013): RTC

=

ln(T r ,c ) − ln(T nr 1,c ) − ln(T nr 2,c )

RTM

=

ln(T r ,m ) − ln(T nr ,m )

RT

=

RTM + RTC

Campante, Chor

Obedience, Routine Tasks, and the Pattern of Specialization

13 / 37

Motivation and Overview Two Stylized Facts A Model of Intergenerational Cultural Transmission

Background on the Data Variables The “Specialization Fact” The “Obedience Fact”

A first look at the routineness measures I

Manufacturing features the highest degree of task routineness, on both Table 1 cognitive and manualTask dimensions. Routineness by Broad Sectors Agriculture, Mining & Construction

Manufacturing

Services

Routine cognitive

4.84 (1.90)

5.87 (0.81)

4.00 (1.57)

Nonroutine cognitive, interactive

2.23 (1.81)

1.44 (0.37)

2.24 (0.90)

Nonroutine cognitive, analytic

3.05 (0.94)

2.97 (0.52)

3.76 (0.82)

Routine manual

3.61 (0.43)

3.98 (0.27)

3.74 (0.56)

Nonroutine manual

2.06 (0.45)

1.32 (0.33)

1.14 (0.76)

Note: The task routineness measures are based on the 1977 DOT coding and are mapped to the industry level using 1960 US Census industry weights, as constructed by Autor, Levy & Murnane (2003). Each index takes values from 0-10. For each

I column, a simplemanufacturing: average of each routineness measure is taken over the industries each broad set sectors, with task the Within Substantial variation inwithin cognitive vsofmanual standard deviation reported in parentheses. For "Agriculture, Mining & Construction", this comprises industries on the Autor, Details Levy & Murnane (2003) Ind6090 CIC codes ranging from 16 to 66; for "Manufacturing", this comprises codes ranging from 100 to 392; and for "Services", this comprises codes ranging from 400 to 901.

routineness.

Campante, Chor

Obedience, Routine Tasks, and the Pattern of Specialization

14 / 37

Motivation and Overview Two Stylized Facts A Model of Intergenerational Cultural Transmission

Background on the Data Variables The “Specialization Fact” The “Obedience Fact”

The “Specialization Fact”: Workplace Obedience and the Pattern of Specialization

Campante, Chor

Obedience, Routine Tasks, and the Pattern of Specialization

15 / 37

Motivation and Overview Two Stylized Facts A Model of Intergenerational Cultural Transmission

Background on the Data Variables The “Specialization Fact” The “Obedience Fact”

Uncovering Specialization Patterns log(Exportcit )

=

α0 + α1 AvgObedWorkc,t−5 × RTi +

X

αlm Ll,c,t−5 × Mm,i

{l,m}

+Dct + Dci + ict I

c: country; i: industry

I

t: 1990-1994, 1995-1999, 2000-2004, 2005-2009, 2010-2014

I

log(Exportcit ): From Feenstra et al. (2005) World Trade Flows dataset (extended to 2014)

I

Uncovering sources of comparative advantage through interaction terms between exporter country characteristics (AvgObedWorkct , Ll,ct ) and industry characteristics (RTi , Mm,i ); c.f., Nunn and Trefler (2014)

Campante, Chor

Obedience, Routine Tasks, and the Pattern of Specialization

(3)

16 / 37

Motivation and Overview Two Stylized Facts A Model of Intergenerational Cultural Transmission

Background on the Data Variables The “Specialization Fact” The “Obedience Fact”

Uncovering Specialization Patterns log(Exportcit )

=

α0 + α1 AvgObedWorkc,t−5 × RTi +

X

αlm Ll,c,t−5 × Mm,i

{l,m}

+Dct + Dci + ict I

Dct : Country-year fixed effects

I

Dci : Country-industry fixed effects

I

Difficulty in comparing obedience scores across countries

(3)

⇒ Focus on how within-country changes in country characteristics affect the pattern of specialization across industries I

OLS; country-clustered standard errors

I

In practice: Use a five-year lagged value of AvgObedWorkct

Campante, Chor

Obedience, Routine Tasks, and the Pattern of Specialization

16 / 37

Motivation and Overview Two Stylized Facts A Model of Intergenerational Cultural Transmission

Background on the Data Variables The “Specialization Fact” The “Obedience Fact”

The Specialization Fact I

Higher workplace obedience associated with ↑ exports in routine industries

I

Even controlling for Heckscher-Ohlin forces (Romalis 2004) and several institutional sources of comparative advantage (rule of law: Nunn 2007, Levchenko 2007, Costinot 2009; financial devt: Manova 2013)

Dependent variable: Routineness measure:

AvgObedWork c,t-5 × Routinenessi

(1) RT

(2) RT

1.6275*** [0.3904]

6.0246*** [1.0643]

Phy. Capital Stockc,t-5 × Capital Intensityi Human Capital Stockc,t-5 × Skill Intensityi

Log (Exportscit) (3) (4) RT RT

(5) RTC

(6) RTM

2.7224*** [1.0070]

2.8511** [1.1516]

2.3146** [1.1393]

1.3896 [2.1153]

0.1852*** [0.0583] 1.2670*** [0.2385]

0.1847*** [0.0599] 1.0783*** [0.2489]

0.1923*** [0.0613] 1.0573*** [0.2643]

0.1958*** [0.0596] 1.2337*** [0.2376]

Rule of Lawc,t-5 × Industryi dummies? Financial Devtc,t-5 × Industryi dummies?

N N

N N

N N

Y Y

Y Y

Y Y

Industries Country-year (ct) dummies? Country-industry (ci) dummies?

All Y Y

Mfg Y Y

Mfg Y Y

Mfg Y Y

Mfg Y Y

Mfg Y Y

19,589 58 0.9500

17,063 58 0.9523

16,194 58 0.9579

15,016 56 0.9611

15,016 56 0.9611

15,016 56 0.9610

Observations No. of countries R2

Campante, Chor

Obedience, Routine Tasks, and the Pattern of Specialization

17 / 37

Motivation and Overview Two Stylized Facts A Model of Intergenerational Cultural Transmission

Background on the Data Variables The “Specialization Fact” The “Obedience Fact”

The Specialization Fact: Remarks I

Effects load on RTC (cognitive) rather than RTM (manual)

I

Coefficient estimate of 2.3146 for AvgObedWorkc,t−5 × RTCi : I

For the median five-year change in AvgObedWorkc,t−5 , exports 0.7% lower in a one s.d. higher RTCi industry.

I

Comparable order of magnitude, but smaller, than corresponding Heckscher-Ohlin effects (3.0% for physical capital; 4.6% for human capital)

Campante, Chor

Obedience, Routine Tasks, and the Pattern of Specialization

18 / 37

Motivation and Overview Two Stylized Facts A Model of Intergenerational Cultural Transmission

Background on the Data Variables The “Specialization Fact” The “Obedience Fact”

The Specialization Fact: Remarks I

Effects load on RTC (cognitive) rather than RTM (manual)

I

Coefficient estimate of 2.3146 for AvgObedWorkc,t−5 × RTCi :

I

I

For the median five-year change in AvgObedWorkc,t−5 , exports 0.7% lower in a one s.d. higher RTCi industry.

I

Comparable order of magnitude, but smaller, than corresponding Heckscher-Ohlin effects (3.0% for physical capital; 4.6% for human capital)

Robust to:

More

I

AvgObedWorkc,t−5 interacted with industry skill-intensity

I

Country human capital interacted with RTCi

I

Interaction terms involving other closely-related country cultural characteristics and RTCi : Importance of independence; Hard work; Work as a duty to society; WVS Questions Individualism; Obedience in children. Campante, Chor

Obedience, Routine Tasks, and the Pattern of Specialization

18 / 37

Motivation and Overview Two Stylized Facts A Model of Intergenerational Cultural Transmission

Background on the Data Variables The “Specialization Fact” The “Obedience Fact”

Table 3 Fact": Controlling for Other Cultural Variables The Specialization Fact:"Specialization Other Cultural Traits Dependent variable: Routineness measure:

(1) RTC

(2) RTC

(3) RTC

Log (Exportscit) (4) RTC

(5) RTC

(6) RTC

(7) RTC

AvgObedWorkc,t-5 × Routinenessi

3.9917** [1.6524]

3.9970** [1.6693]

3.9946** [1.7233]

2.6886* [1.4644]

6.1064*** [1.9315]

4.0555** [1.7061]

4.9030*** [1.6655]

Phy. Capital Stockc,t-5 × Capital Intensityi

0.1693*** [0.0606] 0.6233* [0.3556]

0.1693*** [0.0607] 0.6233* [0.3557]

0.1613** [0.0607] 0.6214* [0.3555]

0.1465** [0.0611] 1.0186** [0.3907]

0.1118** [0.0533] 0.7897* [0.3997]

0.1683*** [0.0605] 0.6228* [0.3556]

0.1221** [0.0549] 1.0780** [0.3974]

-0.6398** [0.2857] 3.2687 [1.9568]

-0.6414** [0.2931] 3.2688 [1.9574]

-0.5343* [0.3115] 3.2677 [1.9569]

-0.2534 [0.3496] 3.0911 [2.0268]

-0.5444 [0.3654] 5.9223*** [1.7641]

-0.6059* [0.3040] 3.2682 [1.9569]

0.1140 [0.3833] 5.0543*** [1.7329]

Human Capital Stockc,t-5 × Skill Intensityi Human Capital Stockc,t-5 × Routinenessi AvgObedWorkc,t-5 × Skill Intensityi Indepedencec,t-5 × Routinenessi

0.0866 [2.0325]

-1.5092 [2.9918]

4.5815** [2.2308] 0.2240 [2.3170] 1.9561** [0.7413] -0.6223 [0.9552] -3.0561 [3.0319]

Y Y

Y Y

Y Y

Y Y

Y Y

Y Y

Y Y

Mfg Y Y

Mfg Y Y

Mfg Y Y

Mfg Y Y

Mfg Y Y

Mfg Y Y

Mfg Y Y

15,016 56 0.9612

15,016 56 0.9612

15,016 56 0.9612

11,256 41 0.9661

10,842 40 0.9687

15,016 56 0.9612

9,669 35 0.9712

Hard Workc,t-5 × Routinenessi

3.2748** [1.4907]

Work as a Dutyc,t-5 × Routinenessi

0.9121 [0.6829]

Individualismc,t-5 × Routinenessi

-0.1477 [0.8867]

AvgObedChildrenc,t-5 × Routinenessi Rule of Lawc,t-5 × Industryi dummies? Financial Devtc,t-5 × Industryi dummies? Industries Country-year (ct) dummies? Country-industry (ci) dummies? Observations No. of countries R2

Notes: Standard errors are clustered by country; ***, **, and * denote significance at the 1%, 5%, and 10% levels respectively. The dependent variable is log exports at the country-industry level averaged over each five-year window (1990-1994 through 2010-2014), where the industry classification follows the Ind6090 CIC codes from Autor, Levy and Murnane (2003). In all

Campante, Chor

Obedience, Routine Tasks, and the Pattern of Specialization

19 / 37

Motivation and Overview Two Stylized Facts A Model of Intergenerational Cultural Transmission

Background on the Data Variables The “Specialization Fact” The “Obedience Fact”

The Specialization Fact: Additional Specifications I

Effects stem from within-countryTable variation: significance weakens when 4 Fact": replacing Dci with Di ;"Specialization see Column 4 Additional Specifications (2) RTC Lag Dep Var

Log (Exportscit) (3) RTC Arellano-Bond

Dependent variable: Routineness measure:

(1) RTC ObedWork (IV)

(4) RTC Other FEs

(5) RTC Other FEs

AvgObedWorkc,t-5 × Routinenessi

3.6734** [1.5079]

3.2397*** [1.0380]

17.6514** [8.0174]

0.1694 [0.4881]

3.7812** [1.5760]

Phy. Capital Stockc,t-5 × Capital Intensityi

0.1682*** [0.0601] 0.7359* [0.3831]

0.1305*** [0.0443] 0.2081 [0.2524]

-0.1038 [0.2537] 0.2925 [0.6438]

0.2494*** [0.0818] 0.5247** [0.2012]

0.0406 [0.0818] -0.2826 [0.6041]

-0.5047 [0.3181] 3.0168* [1.7696]

-0.8325*** [0.2039] 2.9458** [1.2398]

-0.2859 [0.6767] 1.8218 [9.0203]

0.2254 [0.1820] 0.0079 [0.5112]

-0.1287 [0.5703] 4.2063** [1.9155]

0.2989*** [0.0243]

0.3870*** [0.0430]

Y Y

Y Y

Y Y

Y Y Mfg Cty-Yr; Cty-Ind; Ind-Yr

Human Capital Stockc,t-5 × Skill Intensityi Human Capital Stockc,t-5 × Routinenessi AvgObedWorkc,t-5 × Skill Intensityi Log (Exportsci,t-5) Rule of Lawc,t-5 × Industryi dummies? Financial Devtc,t-5 × Industryi dummies? Industries Fixed effects Observations No. of countries R2 F-stat AR1 AR2 Sargan

Y Y Mfg

Mfg

Mfg

Mfg

Cty-Yr; Cty-Ind

Cty-Yr; Cty-Ind

Cty-Yr; Cty-Ind

Cty-Yr; Ind

15,016 56 0.9611 93.17 -------

14,857 56 0.9670 ---------

11,546 55 ----0.0000 0.2211 372.32

15,016 56 0.8169 ---------

15,016 56 0.9641 ---------

Notes: Standard errors are clustered by country; ***, **, andCampante, * denote significance 5%, and 10% Routine levels respectively. dependent variableof is log exports at the Chor at the 1%,Obedience, Tasks,The and the Pattern Specialization

20 / 37

Motivation and Overview Two Stylized Facts A Model of Intergenerational Cultural Transmission

Background on the Data Variables The “Specialization Fact” The “Obedience Fact”

The “Obedience Fact”: From the (lagged) Pattern of Specialization to Workplace Obedience

Campante, Chor

Obedience, Routine Tasks, and the Pattern of Specialization

21 / 37

Motivation and Overview Two Stylized Facts A Model of Intergenerational Cultural Transmission

Background on the Data Variables The “Specialization Fact” The “Obedience Fact”

Obedience Fact: Specification Qn: How does past exposure to a routine task-intensive economic environment shape one’s attitudes towards obedience? I

Start from the negative correlation between education and obedience/ conformity (From sociology, argued to be causal: Kohn 1977; Bowles and Gintis 2001)

I

Quote

Explore how the strength of this effect of education is influenced by growing up in a routine task-intensive economy

Campante, Chor

Obedience, Routine Tasks, and the Pattern of Specialization

22 / 37

Motivation and Overview Two Stylized Facts A Model of Intergenerational Cultural Transmission

Background on the Data Variables The “Specialization Fact” The “Obedience Fact”

Obedience Fact: Specification Qn: How does past exposure to a routine task-intensive economic environment shape one’s attitudes towards obedience? ObedWorkr ,cbw

=

β0 + β1 Educr ,cbw + β2 Educr ,cbw ×expRTCageAcb +βX Xr ,cbw + Dcbg + Dcw + r ,cbw

I

expRTC : Compute weighted-average RTC of country exports from 1962 through 2014 (Feenstra et al.)

I

expRTCageAcb : Export-routineness that birth cohort b in country c was exposed to when they were age A, where A = 0, 5, 10 . . .

(4)

For e.g.: expRTCage10cb for the cohort born in the years b = 1975-1979 is the value of expRTC for 1985-1989. I

(Take five-year window averages. For transition countries, associate the expRTC values of the original country at the time of cohort exposure.)

Campante, Chor

Obedience, Routine Tasks, and the Pattern of Specialization

22 / 37

Motivation and Overview Two Stylized Facts A Model of Intergenerational Cultural Transmission

Background on the Data Variables The “Specialization Fact” The “Obedience Fact”

Obedience Fact: Specification Qn: How does past exposure to a routine task-intensive economic environment shape one’s attitudes towards obedience? ObedWorkr ,cbw

=

β0 + β1 Educr ,cbw + β2 Educr ,cbw ×expRTCageAcb +βX Xr ,cbw + Dcbg + Dcw + r ,cbw

I

(5)

A stringent specification, with β2 estimated from. . . I

within-country-wave, cross-cohort variation in expRTC exposure, and

I

within-country-cohort-gender variation across individuals with different levels of education.

Campante, Chor

Obedience, Routine Tasks, and the Pattern of Specialization

22 / 37

Motivation and Overview Two Stylized Facts A Model of Intergenerational Cultural Transmission

Background on the Data Variables The “Specialization Fact” The “Obedience Fact”

Export-Routineness

1

1.5

1: USA

-.5

0

.5

expRTC expRTM expRT

3: GBR

-.5

0

.5

1

1.5

2: CHN

1960

1980

2000

20201960

1980

2000

2020

year Graphs by gpnam

Campante, Chor

Obedience, Routine Tasks, and the Pattern of Specialization

23 / 37

Motivation and Overview Two Stylized Facts A Model of Intergenerational Cultural Transmission

Background on the Data Variables The “Specialization Fact” The “Obedience Fact”

Instrumental Variable Idea: Exploit a source of variation in country export profiles that is plausibly driven by external conditions. For each five-year window starting in year t, predict country c exports in industry i using: ^ ci,t+s = Exportsci,t−1 × Exports−c,i,t+s . Exports Exports−c,i,t−1

(6)

I

‘−c ’: World trade excluding exports and imports involving c

I

^ i , using above predicted exports as weights Construct the IV, expRTC

I

In principle, predicted export mix should reflect forces related to world export demand, or broader technological forces that affect global supply,. . . . . . rather than policy or socioeconomic shifts specific to country c.

Campante, Chor

Obedience, Routine Tasks, and the Pattern of Specialization

24 / 37

Motivation and Overview Two Stylized Facts A Model of Intergenerational Cultural Transmission

Background on the Data Variables The “Specialization Fact” The “Obedience Fact”

Obedience Fact: Baseline I

More educated individuals less likely to “just follow instructions”

I

BUT: Greater exposure to cognitive export-routineness during one’s schooling years weakens this negative effect of education on obedience.

I

(Holds in both OLS and IV.) Dependent variable:

Importance of obedience in the workplace (3) (4) (5) (6) Age 10 Age 15 Age 20 Age 25

(1) Age 0

(2) Age 5

-0.0198*** [0.0049]

-0.0204*** [0.0046]

-0.0207*** [0.0042]

-0.0220*** [0.0036]

-0.0239*** [0.0033]

Educr ×ExpRTCexposurecb

0.0110 [0.0077]

0.0143** [0.0068]

0.0159*** [0.0057]

0.0111* [0.0062]

0.0078 [0.0058]

0.0053 [0.0066]

0.0011 [0.0064]

-0.0002 [0.0069]

Observations No. of countries R2

50,497 65 0.0727

65,199 65 0.0719

78,809 65 0.0740

90,112 65 0.0740

99,228 65 0.0756

106,399 65 0.0775

112,779 65 0.0785

111,169 65 0.0792

Y Y

Y Y

Y Y

Y Y

ExpRTC exposure at:

(7) Age 30

(8) Age 35

-0.0250*** [0.0031]

-0.0250*** [0.0030]

A: OLS

Educationr

Additional controls: Country-wave (cw) dummies? Cty-cohort-gender (cbg) dummies?

Campante, Chor

-0.0247*** [0.0033]

All columns: Number of children, Marital status Y Y Y Y Y Y Y Y

Obedience, Routine Tasks, and the Pattern of Specialization

25 / 37

Motivation and Overview Two Stylized Facts A Model of Intergenerational Cultural Transmission

Background on the Data Variables The “Specialization Fact” The “Obedience Fact”

Obedience Fact: Baseline I

More educated individuals less likely to “just follow instructions”

I

BUT: Greater exposure to cognitive export-routineness during one’s schooling years weakens this negative effect of education on obedience.

I

(Holds in both OLS and IV.) Dependent variable: ExpRTC exposure at:

(1) Age 0

(2) Age 5

Importance of obedience in the workplace (3) (4) (5) (6) Age 10 Age 15 Age 20 Age 25

(7) Age 30

(8) Age 35

B: Instrumental Variables

Educationr

-0.0208*** [0.0048]

-0.0199*** [0.0048]

-0.0207*** [0.0043]

-0.0215*** [0.0038]

-0.0241*** [0.0034]

-0.0250*** [0.0032]

-0.0256*** [0.0031]

-0.0253*** [0.0030]

Educr ×ExpRTCexposurecb

0.0080 [0.0084]

0.0160** [0.0073]

0.0165*** [0.0061]

0.0151** [0.0061]

0.0086 [0.0066]

0.0057 [0.0070]

0.0023 [0.0067]

0.0008 [0.0070]

Observations No. of countries R2 Kleinberger-Paap Wald F-stat

49,907 65 0.0725 419.47

64,717 65 0.0718 642.49

77,572 65 0.0737 774.09

87,685 65 0.0729 1097.48

96,616 65 0.0752 874.51

103,494 65 0.0775 960.49

109,552 65 0.0783 871.03

108,111 65 0.0785 745.28

Y Y

Y Y

Y Y

Y Y

Additional controls: Country-wave (cw) dummies? Cty-cohort-gender (cbg) dummies?

Campante, Chor

All columns: Number of children, Marital status Y Y Y Y Y Y Y Y

Obedience, Routine Tasks, and the Pattern of Specialization

25 / 37

Motivation and Overview Two Stylized Facts A Model of Intergenerational Cultural Transmission

Background on the Data Variables The “Specialization Fact” The “Obedience Fact”

Obedience Fact: Baseline I

Effects strongest for age 5 and age 10 exposure. Absent for exposure after age 20.

I

Consistent with a story of cultural transmission at a young age. (“impressionable years”; e.g., Giuliano and Spilimbergo 2014) Dependent variable: ExpRTC exposure at:

(1) Age 0

(2) Age 5

Importance of obedience in the workplace (3) (4) (5) (6) Age 10 Age 15 Age 20 Age 25

(7) Age 30

(8) Age 35

B: Instrumental Variables

Educationr

-0.0208*** [0.0048]

-0.0199*** [0.0048]

-0.0207*** [0.0043]

-0.0215*** [0.0038]

-0.0241*** [0.0034]

-0.0250*** [0.0032]

-0.0256*** [0.0031]

-0.0253*** [0.0030]

Educr ×ExpRTCexposurecb

0.0080 [0.0084]

0.0160** [0.0073]

0.0165*** [0.0061]

0.0151** [0.0061]

0.0086 [0.0066]

0.0057 [0.0070]

0.0023 [0.0067]

0.0008 [0.0070]

Observations No. of countries R2 Kleinberger-Paap Wald F-stat

49,907 65 0.0725 419.47

64,717 65 0.0718 642.49

77,572 65 0.0737 774.09

87,685 65 0.0729 1097.48

96,616 65 0.0752 874.51

103,494 65 0.0775 960.49

109,552 65 0.0783 871.03

108,111 65 0.0785 745.28

Y Y

Y Y

Y Y

Y Y

Additional controls: Country-wave (cw) dummies? Cty-cohort-gender (cbg) dummies?

Campante, Chor

All columns: Number of children, Marital status Y Y Y Y Y Y Y Y

Obedience, Routine Tasks, and the Pattern of Specialization

26 / 37

Motivation and Overview Two Stylized Facts A Model of Intergenerational Cultural Transmission

Background on the Data Variables The “Specialization Fact” The “Obedience Fact”

Obedience Fact: With Educr ,cbw × Dcw interactions I

Soaks up the effect that contemporaneous country conditions may have on the education coefficient.

I

Magnitudes: For age 5, one s.d. increase in expRTC translates into a 0.11 s.d. increase in ObedWork for an individual with Educ = 8 (complete tertiary). Cumulative effect potentially bigger if individuals are exposed to a persistent increase in expRTC . Dependent variable:

Importance of obedience in the workplace (3) (4) (5) (6) Age 10 Age 15 Age 20 Age 25

(1) Age 0

(2) Age 5

Educationr

0.0187*** [0.0041]

0.0230*** [0.0027]

0.0088*** [0.0025]

0.0101*** [0.0017]

0.0070*** [0.0010]

Educr ×ExpRTCexposurecb

0.0222** [0.0104]

0.0346*** [0.0077]

0.0196** [0.0077]

0.0177*** [0.0062]

0.0064 [0.0071]

0.0014 [0.0060]

-0.0024 [0.0069]

-0.0042 [0.0078]

49,907 65 0.0771 87.92

64,717 65 0.0760 153.39

77,572 65 0.0777 144.45

87,685 65 0.0768 195.42

96,616 65 0.0788 166.76

103,494 65 0.0811 217.45

109,552 65 0.0816 227.32

108,111 65 0.0820 239.66

Y Y Y

Y Y Y

Y Y Y

Y Y Y

ExpRTC exposure at:

(7) Age 30

(8) Age 35

0.0015** [0.0006]

0.0011 [0.0013]

B: Instrumental Variables

Observations No. of countries R2 Kleinberger-Paap Wald F-stat Additional controls: Country-wave (cw) dummies? Cty-cohort-gender (cbg) dummies? Educr ×Country-wave (cw) dummies?

Campante, Chor

0.0031*** [0.0004]

All columns: Number of children, Marital status Y Y Y Y Y Y Y Y Y Y Y Y

Obedience, Routine Tasks, and the Pattern of Specialization

27 / 37

Motivation and Overview Two Stylized Facts A Model of Intergenerational Cultural Transmission

Background on the Data Variables The “Specialization Fact” The “Obedience Fact”

Additional remarks Results are specific to obedience in the workplace: I

No distinct pattern when looking at the importance of obedience as a Table quality in children

Results robust to . . . I

Dropping countries with ≥ 5% of world exports or imports

I

Controlling for Educr ,cbw interacted with country-cohort exposure (at age A) to the skill-intensity, as well as the capital-intensity, of exports

I

Controlling for Educr ,cbw interacted with country-cohort exposure (at age A) to openness, income per capita, democracy

I

Using overall routineness (RT ) instead of RTC

I

Using the cognitive routineness of manufacturing exports

I

Dropping transition countries Campante, Chor

Table

Table

Obedience, Routine Tasks, and the Pattern of Specialization

28 / 37

Motivation and Overview Two Stylized Facts A Model of Intergenerational Cultural Transmission

Setup Discussion of results “Obedience Traps”

A Model of Intergenerational Cultural Transmission

Campante, Chor

Obedience, Routine Tasks, and the Pattern of Specialization

29 / 37

Motivation and Overview Two Stylized Facts A Model of Intergenerational Cultural Transmission

Setup Discussion of results “Obedience Traps”

Overview of Model Qn: What are the long-term implications of these two empirical Facts regarding a culture of workplace obedience? Develop: A model to capture the interplay between specialization patterns and workplace obedience. 0. Two sectors: a “Basic” sector where workplace obedience raises productivity, and a “Complex” sector where the converse holds 1. Obedience and Human capital at time t determine the pattern of specialization at time t 2. Specialization patterns at time t in turn affect Obedience and Human capital at time t + 1, through the endogenous decisions that parents make over cultural transmission and schooling investments respectively (a la Bisin-Verdier)

Campante, Chor

Obedience, Routine Tasks, and the Pattern of Specialization

30 / 37

Motivation and Overview Two Stylized Facts A Model of Intergenerational Cultural Transmission

Setup Discussion of results “Obedience Traps”

Setup Consider a representative individual (r ) with an endowment of human capital, Hrt , and workplace obedience, θrt , at time t. I

Decides how to allocate Hrt across production activities (hBrt + hCrt = Hrt )

I

B (“Basic”): Routine sector where θrt is complementary to human capital yBrt = AB (f (θrt )hBrt )β Assume: f 0 > 0, f 00 ≤ 0, and 0 < β < 1.

Campante, Chor

Obedience, Routine Tasks, and the Pattern of Specialization

31 / 37

Motivation and Overview Two Stylized Facts A Model of Intergenerational Cultural Transmission

Setup Discussion of results “Obedience Traps”

Setup Consider a representative individual (r ) with an endowment of human capital, Hrt , and workplace obedience, θrt , at time t. I

Decides how to allocate Hrt across production activities (hBrt + hCrt = Hrt )

I

B (“Basic”): Routine sector where θrt is complementary to human capital yBrt = AB (f (θrt )hBrt )β Assume: f 0 > 0, f 00 ≤ 0, and 0 < β < 1.

I

C (“Complex”): Nonroutine sector where θrt hurts the productivity of human capital, and where the nature of nonroutine activity generates scope for human capital externalities yCrt = AC (g (θrt )hCrt )γ

1−γ

Z g (θ˜r t )hC˜r t ˜ r ∈R

0

00

Assume: g < 0, g ≤ 0, and 0 < γ < 1. Campante, Chor

Obedience, Routine Tasks, and the Pattern of Specialization

31 / 37

Motivation and Overview Two Stylized Facts A Model of Intergenerational Cultural Transmission

Setup Discussion of results “Obedience Traps”

Setup I

Overlapping generations: Individuals maximize the expected discounted value of income earned by herself and her (unique) descendent max

hBrt ,hCrt ,Hr ,t+1 ,θr ,t+1

yBrt + pCt yCrt + δEt (yBr ,t+1 + pC ,t+1 yCr ,t+1 ) −ω(Hr ,t+1 ) − τ (θr ,t+1 − θrt )

I

Choice variables related to the next generation: I

Hr ,t+1 : Human capital. Cost in monetary terms given by ω(Hr ,t+1 ), where ω 0 > 0 and ω 00 > 0.

I

θr ,t+1 : Attitudes towards obedience instilled. “Inertia” cost given by τ (θr ,t+1 − θrt ), where τ 0 (0) = 0 and τ 00 > 0.

I

(Assume: Et (AB,t+1 ) = ABt , Et (AC ,t+1 ) = ACt , Et (pC ,t+1 ) = pCt .)

Campante, Chor

Obedience, Routine Tasks, and the Pattern of Specialization

32 / 37

Motivation and Overview Two Stylized Facts A Model of Intergenerational Cultural Transmission

Setup Discussion of results “Obedience Traps”

Characterization Case I: Complete specialization in the B-sector ( pCAAB C large) I

Individual decisions push economy towards raising both θ and H.

I

Any shift in the deep parameters of the model – in particular, tends to raise θ will also raise H in steady state.

AB pC AC

– that

⇒ A complementarity between human capital and obedience, when only the B sector is operative. I

Rationalizing early stages of development? I

East Asia (high θ, high H) vs Latin America (low θ, low H)

Campante, Chor

Obedience, Routine Tasks, and the Pattern of Specialization

33 / 37

Motivation and Overview Two Stylized Facts A Model of Intergenerational Cultural Transmission

Setup Discussion of results “Obedience Traps”

Characterization (cont.) Case II: Diversified economy. I

Define ρt ≡ yBt /yCt to be the “routineness” of the economy at time t.

I

FOC with respect to θt+1 implies: ρ affects the transmission of pro-obedience attitudes, in a manner consistent with the Obedience Fact. (i) When the economy is very nonroutine: (ii) Conversely, when it is very routine: (iii)

I

∂θ ∂H

∂θ ∂H

∂θ ∂H

< 0 in a neighborhood of ρ = 0

> 0 as ρ −→ ∞

is increasing in ρ

Combined with the FOC with respect to Ht+1 , one can then break the complementarity between obedience and human capital.

Campante, Chor

Obedience, Routine Tasks, and the Pattern of Specialization

34 / 37

Motivation and Overview Two Stylized Facts A Model of Intergenerational Cultural Transmission

Setup Discussion of results “Obedience Traps”

Obedience Traps When the economy is diversified (i.e., Case II): I

Steady state in θ and H pinned down by the two FOCs

I

Multiple steady states are possible, especially given the presence of human capital externalities in the C sector I

I

If p AAB lies in an intermediate range of values, get two stable steady states: C C (i) θ = 1; and (ii) θ low (Also, one unstable steady state in between.)

Campante, Chor

Figure

Obedience, Routine Tasks, and the Pattern of Specialization

35 / 37

Motivation and Overview Two Stylized Facts A Model of Intergenerational Cultural Transmission

Setup Discussion of results “Obedience Traps”

Obedience Traps When the economy is diversified (i.e., Case II): I

Steady state in θ and H pinned down by the two FOCs

I

Multiple steady states are possible, especially given the presence of human capital externalities in the C sector I

I

I

If p AAB lies in an intermediate range of values, get two stable steady states: C C (i) θ = 1; and (ii) θ low (Also, one unstable steady state in between.)

Figure

Upshot: Can end up in a high-θ “obedience trap”: I

The predominant workplace mindset is to follow instructions, and the economy is tilted towards the routine B-sector. . .

I

. . . at the expense of the development and expansion of the nonroutine C -sector.

Campante, Chor

Obedience, Routine Tasks, and the Pattern of Specialization

35 / 37

Motivation and Overview Two Stylized Facts A Model of Intergenerational Cultural Transmission

Setup Discussion of results “Obedience Traps”

Obedience Traps When the economy is diversified (i.e., Case II): I

Steady state in θ and H pinned down by the two FOCs

I

Multiple steady states are possible, especially given the presence of human capital externalities in the C sector I

I

If p AAB lies in an intermediate range of values, get two stable steady states: C C (i) θ = 1; and (ii) θ low (Also, one unstable steady state in between.)

Figure

I

Potentially: A middle-income developmental trap.

I

Numerical examples point to aggregate output at the high-θ steady state often being smaller than that in the low-θ steady state (especially so the smaller is β)

Campante, Chor

Obedience, Routine Tasks, and the Pattern of Specialization

35 / 37

Motivation and Overview Two Stylized Facts A Model of Intergenerational Cultural Transmission

Setup Discussion of results “Obedience Traps”

Obedience Traps When the economy is diversified (i.e., Case II): I

Steady state in θ and H pinned down by the two FOCs

I

Multiple steady states are possible, especially given the presence of human capital externalities in the C sector I

I

I

If p AAB lies in an intermediate range of values, get two stable steady states: C C (i) θ = 1; and (ii) θ low (Also, one unstable steady state in between.)

Figure

Moving out of the “obedience trap”? I

Exogenous forces: E.g., a productivity shock that raises AC /AB

I

Policy: E.g., import protection that raises pC

Campante, Chor

Obedience, Routine Tasks, and the Pattern of Specialization

35 / 37

Motivation and Overview Two Stylized Facts A Model of Intergenerational Cultural Transmission

Conclusion

Campante, Chor

Obedience, Routine Tasks, and the Pattern of Specialization

36 / 37

Motivation and Overview Two Stylized Facts A Model of Intergenerational Cultural Transmission

Conclusion I

Report two new facts on the relationship between cultural attitudes towards workplace obedience and the structure of the economy: 1. “Specialization”: Pro-obedience attitudes associated with more exporting in routine industries 2. “Obedience”: In turn, exposure to a more (cognitive) export-routine economy during one’s schooling years shapes pro-obedience attitudes.

I

These inform thinking about a model in which parental decisions are actively made over investment in schooling and the transmission of cultural attitudes: I

How do these co-evolve?

I

How do they shape the structure of the economy in the long run?

Campante, Chor

Obedience, Routine Tasks, and the Pattern of Specialization

37 / 37

Supplementary Slides

Campante, Chor

Obedience, Routine Tasks, and the Pattern of Specialization

1 / 19

More anecdotes: Routine vs nonroutine tasks

Return

“Singaporeans are academically brilliant and they have a tremendous respect for authority. A similar team in the US would keep questioning and want to have a healthy dialogue every step of the way. This may be good in the early stage of a project’s development. But it’s a real problem during the execution. Singaporeans rarely revisit and question the purpose of a task. They have a great ability to translate something from requirement to developed product. They just get it done. . . . [However,] ideas are seldom generated, as no incentives for creativity exist in the Singaporean education system. In three years of operation, our facility has not produced a single patent, and there is no record of new ideas.” (quoting a director of R&D at a medical device MNC located in Singapore)

http://sudhirtv.com/2013/05/17/why-has-singapore-failed-to-prepare-its-citizens-adequately-for-the-knowledge-economy/

Campante, Chor

Obedience, Routine Tasks, and the Pattern of Specialization

2 / 19

More anecdotes: “Obedience Trap”

Return

“East Asians, who all share a tradition of strict discipline, respect for the teacher, no talking back to the teacher and rote learning, must make sure that there is this random intellectual search for new technologies and products.” – Lee Kuan Yew, Foreign Affairs, March/April 1994

“Traditional Chinese culture, still influenced significantly by Confucian values such as ‘obedience’ [and] ‘respect for authority’ (. . .), is not naturally compatible with typical entrepreneurial values.” – Financial Times (Jan 31, 2014)

Campante, Chor

Obedience, Routine Tasks, and the Pattern of Specialization

3 / 19

Appendix Table 1 Countries with Workplace Obedience Data

Back

The Workplace Obedience Measure: List of Available Countries/Territories

Albania (ALB)

Germany (DEU)

Algeria (DZA)

Guatemala (GTM)

Puerto Rico (PRI) Romania (ROM)

Argentina (ARG)

Hong Kong (HKG)

Russia (RUS)

Armenia (ARM)

Hungary (HUN)

Saudi Arabia (SAU)

Australia (AUS)

India (IND)

Serbia (SRB)

Azerbaijan (AZE)

Indonesia (IDN)

Singapore (SGP)

Bangladesh (BGD)

Iran (IRN)

Slovakia (SVK)

Belarus (BLR)

Japan (JPN)

Slovenia (SVN)

Bosnia-Herzegovina (BIH)

Jordan (JOR)

South Africa (ZAF)

Brazil (BRA)

Korea (KOR)

Spain (ESP)

Bulgaria (BGR)

Kyrgyzstan (KGZ)

Sweden (SWE)

Canada (CAN)

Latvia (LVA)

Switzerland (CHE)

Chile (CHL)

Lithuania (LTU)

Taiwan (TWN)

China (CHN)

Macedonia (MKD)

Tanzania (TZA)

Croatia (HRV)

Mexico (MEX)

Turkey (TUR)

Czech Republic (CZE)

Moldova (MDA)

Uganda (UGA)

Dominican Republic (DOM)

Morocco (MAR)

Ukraine (UKR)

Egypt (EGY)

New Zealand (NZL)

United States (USA)

El Salvador (SLV)

Nigeria (NGA)

Uruguay (URY)

Estonia (EST)

Norway (NOR)

Venezuela (VEN)

Finland (FIN)

Peru (PER)

Vietnam (VNM)

Georgia (GEO)

Philippines (PHL)

Zimbabwe (ZWE)

Notes: List of 66 countries/territories in which WVS question C061 on following instructions in the workplace was asked in at least one survey-wave. The "Specialization Fact" regressions in Tables 2-4 and Appendix Tables 5-6 contain fewer countries due to the lack of information on physical and human capital endowments for a small number of countries. The "Obedience Campante, Chor Obedience, Routine Tasks, and the Pattern of Specialization Fact" regressions in Tables 5-7 and Appendix Tables 7-9 contain one fewer country, as there is no information on respondent

4 / 19

Constructing AvgObedWorkct

Back Appendix Table 2 Explaining Attitudes Towards Workplace Obedience

Dependent variable:

Following Instructions in the workplace (1-3) (1) (2) (3)

Educationr

-0.0240*** [0.0043]

-0.0216*** [0.0033]

-0.0202*** [0.0029]

Genderr (1=Female; 0=Male)

-0.0291*** [0.0099]

---

---

N N

Y N

Y Y

Country-cohort-gender (cbg) dummies? Country-wave (cw) dummies?

Dummies for number of children? Dummies for marital status? Dummies for size of town Dummies for employment status? Dummies for occupation? Observations No. of countries R2

Y Y Y Y Y

Additional controls: with p-value of test of joint significance (0.0315) Y (0.2095) Y (0.2227) (0.0000) Y (0.1886) Y (0.0000) (0.0582) Y (0.1683) Y (0.2135) (0.0803) Y (0.0169) Y (0.0028) (0.0000) Y (0.0000) Y (0.0000) 125,709 65 0.0121

125,625 65 0.0771

125,625 65 0.0856

Notes: Standard errors are clustered by country; ***, **, and * denote significance at the 1%, 5%, and 10% levels respectively. The dependent variable is the response provided to WVS question C061 on one's propensity to follow instructions in the workplace. Column 1 contains only respondent characteristics as explanatory variables, while Column 2 adds country-cohort-gender fixed effects, and Column 3 further adds country-survey wave fixed effects. Each column includes full sets of dummy variables for number of children, marital status, size of town, employment status, and occupation of the respondent. The p-value from a F-test for the joint significance of each of these sets of Campante, Chor Obedience, Routine Tasks, and the Pattern of Specialization

5 / 19

Country Workplace Obedience over Time

Workplace Obedience (CAN) Workplace Obedience -.06-.04-.02 0 .02 .04 .06

Workplace Obedience -.06-.04-.02 0 .02 .04 .06

Workplace Obedience (USA)

Back

1985

1990

1995 2000 year

2005

2010

1985

1995 2000 year

2005

2010

Workplace Obedience (KOR) Workplace Obedience -.06-.04-.02 0 .02 .04 .06

Workplace Obedience -.06-.04-.02 0 .02 .04 .06

Workplace Obedience (CHN)

1990

1985

1990

1995 2000 year

2005

2010

Campante, Chor

1985

1990

1995 2000 year

2005

2010

Obedience, Routine Tasks, and the Pattern of Specialization

6 / 19

Workplace Obedience within Cohorts at Different Ages

Back

Mean Workplace Obedience

2.5 2 1.5 1

Oldest Age observed

3

by Country-Cohort-Gender Bins

1

1.5

2 Youngest Age observed

Campante, Chor

2.5

3

Obedience, Routine Tasks, and the Pattern of Specialization

7 / 19

Appendix Table 1 from ALM (2003) APPENDIX 1: DEFINITIONS Variable

DOT deŽnition

OF

TASK MEASURES

Task interpretation Measure of nonroutine analytic tasks

2. Direction, Control, Planning (DCP)

Adaptability to accepting responsibility for the direction, control, or planning of an activity

Measure of nonroutine interactive tasks

3. Set Limits, Tolerances, or Standards (STS)

Adaptability to situations requiring the precise attainment of set limits, tolerances, or standards Ability to move Žngers, and manipulate small objects with Žngers, rapidly or accurately

Measure of routine cognitive tasks

Ability to move the hand and foot coordinately with each other in accordance with visual stimuli

Measure of nonroutine manual tasks

4. Finger Dexterity (FINGDEX)

5. Eye Hand Foot Coordination (EYEHAND)

Measure of routine manual tasks

1977 DICTIONARY

OF

OCCUPATIONAL TITLES

Example tasks from Handbook for Analyzing Jobs Lowest level: Adds and subtracts 2-digit numbers; performs operations with units such as cup, pint, and quart. Midlevel: Computes discount, interest, proŽt, and loss; inspects at glass and compiles defect data based on samples to determine variances from acceptable quality limits. Highest level: Conducts and oversees analyses of aerodynamic and thermodynamic systems . . . to determine suitability of design for aircraft and missiles. Plans and designs private residences, ofŽce buildings, factories, and other structures; applies principles of accounting to install and maintain operation of general accounting system; conducts prosecution in court proceedings . . . gathers and analyzes evidence, reviews pertinent decisions . . . appears against accused in court of law; commands Žshing vessel crew engaged in catching Žsh and other marine life. Operates a billing machine to transcribe from ofŽce records data; calculates degrees, minutes, and second of latitude and longitude, using standard navigation aids; measures dimensions of bottle, using gauges and micrometers to verify that setup of bottle-making conforms to manufacturing speciŽcations; prepares and veriŽes voter lists from ofŽcial registration records. Mixes and bakes ingredients according to recipes; sews fasteners and decorative trimmings to articles; feeds tungsten Žlament wire coils into machine that mounts them to stems in electric light bulbs; operates tabulating machine that processes data from tabulating cards into printed records; packs agricultural produce such as bulbs, fruits, nuts, eggs, and vegetables for storage or shipment; attaches hands to faces of watches. Lowest level: Tends machine that crimps eyelets, grommets; next level: attends to beef cattle on stock ranch; drives bus to transport passengers; next level: pilots airplane to transport passengers; prunes and treats ornamental and shade trees; highest level: performs gymnastic feats of skill and balance.

Source: U. S. Department of Labor, Manpower Administration, Handbook for Analyzing Jobs (Washington, DC, 1972).

Campante, Chor

Obedience, Routine Tasks, and the Pattern of Specialization

1323

General educational development, mathematics

FROM THE

SKILL CONTENT OF TECHNICAL CHANGE

1. GED Math (MATH)

Back

8 / 19

Appendix Table 3

Five Most and Least Routine Manufacturing Industries Cognitive vs ManualThe task routineness (within manufacturing) Cognitive (RTC)

Return

Manual (RTM) Five Least Routine

Five Least Routine Not specified food industries (122)

-0.869

Logging (230)

-0.039

Drugs (181)

-0.613

Dairy Products (101)

0.428

Guided missiles, space vehicles, and parts. Ordnance, and Aircraft and parts (362)

-0.374

Cement, concrete, and gypsum, and plaster products (251)

0.619

Plastics, synthetics, and resins; Soaps and cosmetics; Agricultural Chemicals; Industrial and miscellaneous chemicals (346)

-0.311

Sawmills, planing mills, and millwork (231)

0.620

Newspaper publishing and printing (171)

-0.305

Beverage (120)

0.774

Five Most Routine

Five Most Routine Logging (230)

1.079

Not specified food industries (122)

1.451

Apparel and accessories, except knit (151)

1.080

Engine and turbines; Construction and material handling machines; metalworking machinery; machinery, except electrical, n.e.c.; etc. (176)

1.474

Footwear, except rubber and plastic (221)

1.141

Drugs (181)

1.519

Yarn, thread, and fabric mills (142)

1.308

Newspaper publishing and printing (171)

1.644

1.410

Printing, publishing, and allied industries except newspapers (172)

1.766

Knitting mills (132)

Notes: Based on the Ind6090 CIC codes from Autor, Levy and Murnane (2003). The table lists the five most and least routine manufacturing Campante, Chor Obedience, Routine Tasks, and industries the Pattern industries, for the cognitive and manual routineness measures respectively. The manufacturing industries are those with of CICSpecialization codes

9 / 19

WVS: Obedience as a quality in children

Return

Question A042: “Here is a list of qualities that children can be encouraged to learn at home. Which, if any, do you consider to be especially important? Please choose up to five. (CODE FIVE ONLY)”

I

Out of a list of up to 11 qualities, including: “good manners”, “independence”, “hard work”, “obedience”, “religious faith”, “thrift”

I

Responses:

I

I

‘1’= Important

I

‘0’= Not mentioned

Available in all six waves.

Campante, Chor

Obedience, Routine Tasks, and the Pattern of Specialization

10 / 19

WVS: Work as a Duty

Return

Question C039: “Do you agree or disagree with the following statements?” “Work is a duty towards society”

I

Responses (recoded): I

‘5’= Strongly agree

I

...

I

‘1’= Strongly disagree

Campante, Chor

Obedience, Routine Tasks, and the Pattern of Specialization

11 / 19

WVS: Individualist Attitudes

Return

Question G023: “People have different views about themselves and how they relate to the world. Using this card, would you tell me how strongly you agree or disagree with each of the following statements about how you see yourself?” (Read out and code one answer for each statement) “I see myself as an autonomous individual.”

I

Responses (recoded): I

‘4’= Strongly agree

I

...

I

‘1’= Strongly disagree

Campante, Chor

Obedience, Routine Tasks, and the Pattern of Specialization

12 / 19

Appendix Table 5 "Specialization Fact" Robustness: Other Country Variables The Specialization Fact: Other Country Variables Return (1) RTC

(2) RTC

Log (Exportscit) (3) RTC

(4) RTC

(5) RTC

AvgObedWorkc,t-5 × Routinenessi

3.9795** [1.6498]

4.0281** [1.6500]

3.9982** [1.6577]

3.8857** [1.6281]

3.7796** [1.6403]

Phy. Capital Stockc,t-5 × Capital Intensityi

0.1665*** [0.0601] 0.6223* [0.3555]

0.1815*** [0.0581] 0.6274* [0.3558]

0.1692*** [0.0606] 0.6233* [0.3556]

0.1736*** [0.0601] 0.6263* [0.3554]

0.1815*** [0.0581] 0.6289* [0.3556]

-0.6165* [0.3363] 3.2695 [1.9566]

-0.7404** [0.3475] 3.2687 [1.9579]

-0.6356** [0.3033] 3.2686 [1.9569]

-0.6030* [0.3038] 3.2745* [1.9559]

-0.7000* [0.3597] 3.2779* [1.9572]

Dependent variable: Routineness measure:

Human Capital Stockc,t-5 × Skill Intensityi Human Capital Stockc,t-5 × Routinenessi AvgObedWorkc,t-5 × Skill Intensityi Log GDP per capitac,t-5 × Routinenessi

-0.0348 [0.1505]

Phy. Capital Stockc,t-5 × Routinenessi

-0.1593 [0.2124] 0.1426 [0.1861] 0.0199 [0.0143]

0.0861 [0.1102]

Median Agec,t-5 × Routinenessi

-0.0015 [0.0203]

Age Shares in Popc,t-5 × Routinenessi?

N

N

N

Y

Y

Rule of Lawc,t-5 × Industryi dummies? Financial Devtc,t-5 × Industryi dummies?

Y Y

Y Y

Y Y

Y Y

Y Y

Mfg Y Y

Mfg Y Y

Mfg Y Y

Mfg Y Y

Mfg Y Y

15,016 56 0.9612

15,016 56 0.9612

15,016 56 0.9612

15,016 56 0.9612

15,016 56 0.9613

Industries Country-year (ct) dummies? Country-industry (ci) dummies? Observations No. of countries R2

Notes: Standard errors are clustered by country; ***, **, and * denote significance at the 1%, 5%, and 10% levels respectively. The dependent variable is log exports at the country-industry level averaged over Campante, each five-yearChor window (1990-1994 throughRoutine 2010-2014), where thethe industry classification follows the Obedience, Tasks, and Pattern of Specialization

13 / 19

Appendix Table 6 "Specialization Fact" Robustness: Other Industry Variables The Specialization Fact: Other Industry Variables Return Dependent variable: Routineness measure:

(1) RTC

(2) RTC

Log (Exportscit) (3) RTC

(4) RTC

(5) RTC

AvgObedWorkc,t-5 × Routinenessi

3.2476* [1.6595]

2.9232** [1.4127]

4.2029** [1.6578]

4.1522** [1.7821]

3.5900** [1.6450]

Phy. Capital Stockc,t-5 × Capital Intensityi

0.1616** [0.0615] 0.6206* [0.3553]

0.1384** [0.0671] 0.6145* [0.3563]

0.1688*** [0.0606] 0.6224* [0.3556]

0.1698*** [0.0606] 0.6239* [0.3556]

0.1382** [0.0672] 0.6153* [0.3561]

-0.6481** [0.2849] 2.2229 [1.9888]

-0.6737** [0.2896] 2.9906 [1.9185]

-0.6406** [0.2853] 2.4988 [1.9185]

-0.6390** [0.2859] 3.4105* [2.0325]

-0.6729** [0.2894] 2.0839 [2.0488]

1.5545 [3.3172]

2.8624 [5.4440] -1.1730 [1.1885] 2.9730 [1.8916] 7.9636 [5.2314]

Human Capital Stockc,t-5 × Skill Intensityi Human Capital Stockc,t-5 × Routinenessi AvgObedWorkc,t-5 × Skill Intensityi AvgObedWorkc,t-5 × Value-added Sharei

4.1244 [2.6869]

AvgObedWorkc,t-5 × Capital Intensityi

-1.3002 [0.9414]

AvgObedWorkc,t-5 × Complexityi

2.4489 [1.6361]

AvgObedWorkc,t-5 × Input Herfindahli Rule of Lawc,t-5 × Industryi dummies? Financial Devtc,t-5 × Industryi dummies? Industries Country-year (ct) dummies? Country-industry (ci) dummies? Observations No. of countries R2

Y Y

Y Y

Y Y

Y Y

Y Y

Mfg Y Y

Mfg Y Y

Mfg Y Y

Mfg Y Y

Mfg Y Y

15,016 56 0.9612

15,016 56 0.9612

15,016 56 0.9612

15,016 56 0.9612

15,016 56 0.9612

Notes: Standard errors are clustered by country; ***, **, and * denote significance at the 1%, 5%, and 10% levels respectively. The dependent variable is Chor Obedience, Routine Tasks, andthethe Pattern of Specialization log exports at the country-industry level averaged overCampante, each five-year window (1990-1994 through 2010-2014), where industry classification follows the

14 / 19

More anecdotes: Education and “Conformity”

Return

“Education is important because self-direction requires more intellectual flexibility and breadth of perspective than does conformity; tolerance of nonconformity, in particular, requires a degree of analytic ability that is difficult to achieve without formal education. But education is not all that is involved. The conformity of people at lower social class levels is in large measure a carry-over from the limitations of their occupational experiences.” – Kohn (1977), p.190

Campante, Chor

Obedience, Routine Tasks, and the Pattern of Specialization

15 / 19

Obedience Fact: Robustness

Return

Table 7 "Obedience Fact": Further Robustness

Dependent variable:

Importance of obedience in the workplace (3) (4) (5) (6) Age 10 Age 15 Age 20 Age 25

(1) Age 0

(2) Age 5

Educr × ExpRTCexposurecb

0.0229** [0.0104]

0.0357*** [0.0077]

Educr × ExpRTCexposurecb

0.0002 [0.0056]

-0.0012 [0.0030]

Educr × ExpRTCexposurecb

0.0180* [0.0106]

Educr × ExpRTCexposurecb

0.0244 [0.0276]

0.0317** [0.0143]

Y Y Y

Y Y Y

ExpRTC exposure at:

(7) Age 30

(8) Age 35

-0.0021 [0.0071]

-0.0042 [0.0080]

0.0044 [0.0028]

0.0047 [0.0033]

Instrumental Variables

A: Drop Countries with >=5% Share of World Exports or Imports 0.0217*** [0.0076]

0.0182*** [0.0063]

0.0077 [0.0070]

0.0022 [0.0060]

B: Alternative RHS variable: Importance of Obedience in Children -0.0004 [0.0024]

0.0009 [0.0024]

-0.0009 [0.0029]

0.0017 [0.0027]

C: Controlling for Educr × Country-Cohort Exposure to Export Skill- and Capital-Intensity 0.0380*** [0.0077]

0.0192** [0.0078]

0.0181** [0.0071]

0.0109 [0.0081]

0.0030 [0.0071]

-0.0056 [0.0080]

-0.0136* [0.0071]

D: Controlling for Educr × Country-Cohort Exposure to Income per capita, Openness, Democracy

Additional controls: Country-wave (cw) dummies? Cty-cohort-gender (cbg) dummies? Educr × Country-wave (cw) dummies?

0.0234*** [0.0089]

0.0322*** [0.0071]

0.0163** [0.0080]

0.0012 [0.0072]

All columns: Dummies for Number of children, Marital status Y Y Y Y Y Y Y Y Y Y Y Y

-0.0097 [0.0087]

-0.0179** [0.0080]

Y Y Y

Y Y Y

Notes: Standard errors are clustered by country; ***, **, and * denote significance at the 1%, 5%, and 10% levels respectively. The dependent variable is the response provided to WVS question C061 on one's propensity to follow instructions in the workplace, except in Panel B where the dependent variable is instead drawn from WVS question A042 on the importance of obedience in children. Each successive column tests for whether the cognitive routineness of exports (expRTC) that the respondent was exposed to in the five-year window where he/she turned age A (where A=0, 5, 10 etc.) affects attitudes towards workplace obedience. All columns include survey country-wave and country-cohort-gender fixed effects, respondent education interacted with country-wave fixed effects, as well as a full set of dummies for the number of children and marital status. All columns report instrumental variables estimates using a constructed "shiftshare" IV for expRTC; the Kleinberger-Paap F-statistics obtained range between 19.85 and 406.63. Only the coefficient on the interaction term between respondent education and expRTC

Campante, Chor

Obedience, Routine Tasks, and the Pattern of Specialization

16 / 19

Obedience Fact: Further Robustness

Return

Appendix Table 7 "Obedience Fact": Further Robustness Checks Dependent variable:

Importance of obedience in the workplace (3) (4) (5) (6) Age 10 Age 15 Age 20 Age 25

(1) Age 0

(2) Age 5

Educr × ExpRTexposurecb

0.0170** [0.0075]

0.0244*** [0.0055]

0.0128** [0.0063]

Educr × ExpRTCexposurecb

0.0228** [0.0106]

0.0353*** [0.0072]

0.0208*** [0.0080]

Educr × ExpRTCexposurecb

-0.0134 [0.0425]

0.0744** [0.0306]

0.0715*** [0.0230]

Y Y Y

Y Y Y

ExpRTC exposure at:

(7) Age 30

(8) Age 35

0.0031 [0.0044]

0.0013 [0.0054]

-0.0018 [0.0060]

0.0047 [0.0059]

-0.0018 [0.0072]

-0.0065 [0.0077]

-0.0311* [0.0160]

-0.0140 [0.0195]

0.0056 [0.0196]

Y Y Y

Y Y Y

Instrumental Variables

A: Overall Export Routineness (RT) 0.0128*** [0.0049]

0.0064 [0.0053]

B: Drop transition countries 0.0191*** [0.0061]

0.0123** [0.0053]

C: Manufacturing Exports Only

Additional controls: Country-wave (cw) dummies? Cty-cohort-gender (cbg) dummies? Educr × Country-wave (cw) dummies?

0.0365* [0.0197]

-0.0167 [0.0140]

All columns: Dummies for Number of children, Marital status Y Y Y Y Y Y Y Y Y Y Y Y

Notes: Standard errors are clustered by country; ***, **, and * denote significance at the 1%, 5%, and 10% levels respectively. The dependent variable is the response provided to WVS question C061 on one's propensity to follow instructions in the workplace. Each successive column tests for whether the cognitive routineness of exports (expRTC) that the respondent was exposed to in the five-year window where he/she turned age A (where A=0, 5, 10 etc.) affects attitudes towards workplace obedience. All columns include survey country-wave and countrycohort-gender fixed effects, respondent education interacted with country-wave fixed effects, as well as a full set of dummies for the number of children and marital status. All columns report instrumental variables estimates using a constructed "shift-share" IV for expRTC; the Kleinberger-Paap F-statistics obtained range between 81.39 and 725.54. Only the coefficient on the interaction term between respondent education and expRTC exposure at age A is reported. In Panel A, the measure of export routineness is based on the overall RT index (instead of RTC). In Panel B, countries that experienced political transitions are dropped. In Panel C, the measure of export routineness (RTC) is based only on manufacturing exports.

Campante, Chor

Obedience, Routine Tasks, and the Pattern of Specialization

17 / 19

Obedience Fact: Robustness ObedWorkr ,cbw

=

Return

β0 + β1 Educr ,cbw + β2 Educr ,cbw ×expRTCageAcb +β3 expRTCageAcb + βX Xr ,cbw +Dbg + Dcw + r ,cbw

Dependent variable:

(7)

Importance of obedience in the workplace (3) (4) (5) (6) Age 10 Age 15 Age 20 Age 25

(1) Age 0

(2) Age 5

-0.0207*** [0.0046]

-0.0196*** [0.0046]

-0.0206*** [0.0041]

-0.0213*** [0.0036]

-0.0238*** [0.0032]

Educr ×ExpRTCexposurecb

0.0070 [0.0085]

0.0172** [0.0069]

0.0169*** [0.0062]

0.0163*** [0.0057]

ExpRTCexposurecb

-0.0050 [0.0789]

-0.0979* [0.0592]

-0.0938* [0.0496]

-0.0660 [0.0466]

Observations No. of countries R2 Kleinberger-Paap Wald F-stat

49,911 65 0.0635 36.02

64,721 65 0.0624 55.89

77,576 65 0.0637 49.93

87,689 65 0.0629 61.62

Additional controls: Country-wave (cw) dummies? Cohort-Gender (bg) dummies?

Y Y

Y Y

ExpRTC exposure at:

(7) Age 30

(8) Age 35

-0.0250*** [0.0031]

-0.0258*** [0.0030]

-0.0256*** [0.0029]

0.0136** [0.0055]

0.0088 [0.0064]

0.0050 [0.0060]

0.0036 [0.0066]

-0.0379 [0.0398]

-0.0529 [0.0444]

-0.0420 [0.0488]

-0.0094 [0.0419]

96,620 65 0.0645 41.75

103,498 65 0.0661 43.63

109,557 65 0.0664 46.85

108,115 65 0.0662 45.54

Y Y

Y Y

B: Instrumental Variables

Educationr

All columns: Dummies for Number of children, Marital status Y Y Y Y Y Y Y Y

Campante, Chor

Obedience, Routine Tasks, and the Pattern of Specialization

18 / 19

Obedience Fact: Robustness ObedWorkr ,cbw

=

Return

β0 + β1 Educr ,cbw + β2 Educr ,cbw ×expRTCageAcb +β3 expRTCageAcb + βX Xr ,cbw +Dbg + Dcw + r ,cbw

Dependent variable: ExpRTC exposure at:

(1) Age 0

(2) Age 5

(7)

Importance of obedience in the workplace (3) (4) (5) (6) Age 10 Age 15 Age 20 Age 25

(7) Age 30

(8) Age 35

B: Instrumental Variables

Educationr

-0.0296*** [0.0063]

-0.0265*** [0.0062]

-0.0260*** [0.0052]

-0.0250*** [0.0045]

-0.0266*** [0.0039]

-0.0274*** [0.0036]

-0.0274*** [0.0036]

-0.0266*** [0.0036]

Educr ×ExpRTCexposurecb

0.0001 [0.0091]

0.0120 [0.0082]

0.0129** [0.0064]

0.0174*** [0.0058]

0.0135** [0.0059]

0.0113* [0.0065]

0.0037 [0.0067]

0.0061 [0.0073]

ExpRTCexposurecb

0.0380 [0.0708]

-0.0218 [0.0507]

-0.0425 [0.0476]

-0.0593 [0.0531]

-0.0321 [0.0444]

-0.0648 [0.0574]

-0.0562 [0.0646]

-0.0039 [0.0463]

Observations No. of countries R2 Kleinberger-Paap Wald F-stat

30,202 39 0.0537 12.93

40,135 42 0.0592 19.32

50,574 46 0.0629 29.17

60,130 56 0.0629 49.11

70,041 62 0.0657 57.49

76,939 62 0.0674 42.50

79,246 62 0.0679 36.81

74,196 62 0.0689 36.02

Additional controls: Country-wave (cw) dummies? Cohort-Gender (bg) dummies?

Y Y

All columns: Dummies for Number of children, Marital status, and Country-Cohort Exposure to log GDP per capita, Exports over GDP, Democracy, at age A Y Y Y Y Y Y Y Y Y Y Y Y

Campante, Chor

Obedience, Routine Tasks, and the Pattern of Specialization

Y Y

18 / 19

Obedience Traps

Return

Z(

0

L

M

H 

Drawn for: f (θ) = θ and g (θ) = 1 − θ

Campante, Chor

Obedience, Routine Tasks, and the Pattern of Specialization

19 / 19