The Biocultural Origins of Human Capital Formation July 4, 2015
The Biocultural Origins of Human Capital Formation
Oded Galor and Marc Klemp
July 4, 2015
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 1 / 72
Research Questions
The evolutionary foundations of human capital formation
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 2 / 72
Have the forces of natural selection favored individuals with a larger predisposition towards child quality?
Have the forces of natural selection contributed to:
Human capital formation?
The demographic transition?
The transition from stagnation to growth?
Research Questions
The evolutionary foundations of human capital formation
Have the forces of natural selection favored individuals with a larger predisposition towards child quality?
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 2 / 72
Have the forces of natural selection contributed to:
Human capital formation?
The demographic transition?
The transition from stagnation to growth?
Research Questions
The evolutionary foundations of human capital formation
Have the forces of natural selection favored individuals with a larger predisposition towards child quality?
Have the forces of natural selection contributed to:
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 2 / 72
Human capital formation?
The demographic transition?
The transition from stagnation to growth?
Research Questions
The evolutionary foundations of human capital formation
Have the forces of natural selection favored individuals with a larger predisposition towards child quality?
Have the forces of natural selection contributed to:
Human capital formation?
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 2 / 72
The demographic transition?
The transition from stagnation to growth?
Research Questions
The evolutionary foundations of human capital formation
Have the forces of natural selection favored individuals with a larger predisposition towards child quality?
Have the forces of natural selection contributed to:
Human capital formation?
The demographic transition?
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 2 / 72
The transition from stagnation to growth?
Research Questions
The evolutionary foundations of human capital formation
Have the forces of natural selection favored individuals with a larger predisposition towards child quality?
Have the forces of natural selection contributed to:
Human capital formation?
The demographic transition?
The transition from stagnation to growth?
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 2 / 72
Existing Theories
The Life-History Theory (Lack, 1954)
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 3 / 72
Reproductive strategies of any organisms have evolved to maximize their long-run reproductive success
Clutch size (e.g., number of eggs) maximizes long-run reproduction
Evolutionary Growth Theory (Galor-Moav, 2002)
Natural Selection is the origin of economic growth
Traits associated with pre-disposition towards child quality were selected, triggering the transition from stagnation to growth
Existing Theories
The Life-History Theory (Lack, 1954)
Reproductive strategies of any organisms have evolved to maximize their long-run reproductive success
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 3 / 72
Clutch size (e.g., number of eggs) maximizes long-run reproduction
Evolutionary Growth Theory (Galor-Moav, 2002)
Natural Selection is the origin of economic growth
Traits associated with pre-disposition towards child quality were selected, triggering the transition from stagnation to growth
Existing Theories
The Life-History Theory (Lack, 1954)
Reproductive strategies of any organisms have evolved to maximize their long-run reproductive success
Clutch size (e.g., number of eggs) maximizes long-run reproduction
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 3 / 72
Evolutionary Growth Theory (Galor-Moav, 2002)
Natural Selection is the origin of economic growth
Traits associated with pre-disposition towards child quality were selected, triggering the transition from stagnation to growth
Existing Theories
The Life-History Theory (Lack, 1954)
Reproductive strategies of any organisms have evolved to maximize their long-run reproductive success
Clutch size (e.g., number of eggs) maximizes long-run reproduction
Evolutionary Growth Theory (Galor-Moav, 2002)
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 3 / 72
Natural Selection is the origin of economic growth
Traits associated with pre-disposition towards child quality were selected, triggering the transition from stagnation to growth
Existing Theories
The Life-History Theory (Lack, 1954)
Reproductive strategies of any organisms have evolved to maximize their long-run reproductive success
Clutch size (e.g., number of eggs) maximizes long-run reproduction
Evolutionary Growth Theory (Galor-Moav, 2002)
Natural Selection is the origin of economic growth
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 3 / 72
Traits associated with pre-disposition towards child quality were selected, triggering the transition from stagnation to growth
Existing Theories
The Life-History Theory (Lack, 1954)
Reproductive strategies of any organisms have evolved to maximize their long-run reproductive success
Clutch size (e.g., number of eggs) maximizes long-run reproduction
Evolutionary Growth Theory (Galor-Moav, 2002)
Natural Selection is the origin of economic growth
Traits associated with pre-disposition towards child quality were selected, triggering the transition from stagnation to growth
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 3 / 72
Fundamental Empirical Questions
Is there an intergenerational trade-off in reproductive success?
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 4 / 72
Is higher fecundity (e.g., higher sperm count) conducive for reproductive success in the long run?
Does a larger number of children generate a larger number of offspring in the long run?
Have the forces of natural selection favored moderately fertile individuals (i.e., the quality strategy) during the Malthusian epoch?
Fundamental Empirical Questions
Is there an intergenerational trade-off in reproductive success?
Is higher fecundity (e.g., higher sperm count) conducive for reproductive success in the long run?
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 4 / 72
Does a larger number of children generate a larger number of offspring in the long run?
Have the forces of natural selection favored moderately fertile individuals (i.e., the quality strategy) during the Malthusian epoch?
Fundamental Empirical Questions
Is there an intergenerational trade-off in reproductive success?
Is higher fecundity (e.g., higher sperm count) conducive for reproductive success in the long run?
Does a larger number of children generate a larger number of offspring in the long run?
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 4 / 72
Have the forces of natural selection favored moderately fertile individuals (i.e., the quality strategy) during the Malthusian epoch?
Fundamental Empirical Questions
Is there an intergenerational trade-off in reproductive success?
Is higher fecundity (e.g., higher sperm count) conducive for reproductive success in the long run?
Does a larger number of children generate a larger number of offspring in the long run?
Have the forces of natural selection favored moderately fertile individuals (i.e., the quality strategy) during the Malthusian epoch?
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 4 / 72
Conditions for the Optimality of Moderate Fertility
Necessary Condition: Quantity-Quality Trade-off
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 5 / 72
Quantity of children come on the account of their quality (survival, education, marriageability)
Lower quality reduces the child’s earning capacity and therefore fertility
Sufficient Condition:
The product of the following elasticities is greater than 1:
The elasticity of child quality with respect to to quantity
The elasticity of grandchildren with respect to child quality
Conditions for the Optimality of Moderate Fertility
Necessary Condition: Quantity-Quality Trade-off
Quantity of children come on the account of their quality (survival, education, marriageability)
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 5 / 72
Lower quality reduces the child’s earning capacity and therefore fertility
Sufficient Condition:
The product of the following elasticities is greater than 1:
The elasticity of child quality with respect to to quantity
The elasticity of grandchildren with respect to child quality
Conditions for the Optimality of Moderate Fertility
Necessary Condition: Quantity-Quality Trade-off
Quantity of children come on the account of their quality (survival, education, marriageability)
Lower quality reduces the child’s earning capacity and therefore fertility
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 5 / 72
Sufficient Condition:
The product of the following elasticities is greater than 1:
The elasticity of child quality with respect to to quantity
The elasticity of grandchildren with respect to child quality
Conditions for the Optimality of Moderate Fertility
Necessary Condition: Quantity-Quality Trade-off
Quantity of children come on the account of their quality (survival, education, marriageability)
Lower quality reduces the child’s earning capacity and therefore fertility
Sufficient Condition:
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 5 / 72
The product of the following elasticities is greater than 1:
The elasticity of child quality with respect to to quantity
The elasticity of grandchildren with respect to child quality
Conditions for the Optimality of Moderate Fertility
Necessary Condition: Quantity-Quality Trade-off
Quantity of children come on the account of their quality (survival, education, marriageability)
Lower quality reduces the child’s earning capacity and therefore fertility
Sufficient Condition:
The product of the following elasticities is greater than 1:
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 5 / 72
The elasticity of child quality with respect to to quantity
The elasticity of grandchildren with respect to child quality
Conditions for the Optimality of Moderate Fertility
Necessary Condition: Quantity-Quality Trade-off
Quantity of children come on the account of their quality (survival, education, marriageability)
Lower quality reduces the child’s earning capacity and therefore fertility
Sufficient Condition:
The product of the following elasticities is greater than 1:
The elasticity of child quality with respect to to quantity
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 5 / 72
The elasticity of grandchildren with respect to child quality
Conditions for the Optimality of Moderate Fertility
Necessary Condition: Quantity-Quality Trade-off
Quantity of children come on the account of their quality (survival, education, marriageability)
Lower quality reduces the child’s earning capacity and therefore fertility
Sufficient Condition:
The product of the following elasticities is greater than 1:
The elasticity of child quality with respect to to quantity
The elasticity of grandchildren with respect to child quality
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 5 / 72
Evidence for a Quantity-Quality Trade-off
Plants
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 6 / 72
Trade-off between the number and size of seeds
Animals
Trade-off between the number of offspring and their size
Humans
Trade-off between the number of children and their human capital
Evidence for a Quantity-Quality Trade-off
Plants
Trade-off between the number and size of seeds
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 6 / 72
Animals
Trade-off between the number of offspring and their size
Humans
Trade-off between the number of children and their human capital
Evidence for a Quantity-Quality Trade-off
Plants
Trade-off between the number and size of seeds
Animals
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 6 / 72
Trade-off between the number of offspring and their size
Humans
Trade-off between the number of children and their human capital
Evidence for a Quantity-Quality Trade-off
Plants
Trade-off between the number and size of seeds
Animals
Trade-off between the number of offspring and their size
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 6 / 72
Humans
Trade-off between the number of children and their human capital
Evidence for a Quantity-Quality Trade-off
Plants
Trade-off between the number and size of seeds
Animals
Trade-off between the number of offspring and their size
Humans
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 6 / 72
Trade-off between the number of children and their human capital
Evidence for a Quantity-Quality Trade-off
Plants
Trade-off between the number and size of seeds
Animals
Trade-off between the number of offspring and their size
Humans
Trade-off between the number of children and their human capital
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 6 / 72
Empirics of Intertemporal Trade-off in Reproductive Success in Humans
Poorly identified
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 7 / 72
Unobserved heterogeneity
Reverse causality
Small samples
Often post-demographic transition populations
No intertemporal trade-off in reproductive success:
Kenyan Kipsigis
(Borgerhoff-Mulder, 2000)
82 men and 64 women in 4 income groups
No trade-off for men and for 3 of the 4 income groups
New Mexican men
(Kaplan et al., 1995)
Post-demographic transition
Empirics of Intertemporal Trade-off in Reproductive Success in Humans
Poorly identified
Unobserved heterogeneity
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 7 / 72
Reverse causality
Small samples
Often post-demographic transition populations
No intertemporal trade-off in reproductive success:
Kenyan Kipsigis
(Borgerhoff-Mulder, 2000)
82 men and 64 women in 4 income groups
No trade-off for men and for 3 of the 4 income groups
New Mexican men
(Kaplan et al., 1995)
Post-demographic transition
Empirics of Intertemporal Trade-off in Reproductive Success in Humans
Poorly identified
Unobserved heterogeneity
Reverse causality
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 7 / 72
Small samples
Often post-demographic transition populations
No intertemporal trade-off in reproductive success:
Kenyan Kipsigis
(Borgerhoff-Mulder, 2000)
82 men and 64 women in 4 income groups
No trade-off for men and for 3 of the 4 income groups
New Mexican men
(Kaplan et al., 1995)
Post-demographic transition
Empirics of Intertemporal Trade-off in Reproductive Success in Humans
Poorly identified
Unobserved heterogeneity
Reverse causality
Small samples
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 7 / 72
Often post-demographic transition populations
No intertemporal trade-off in reproductive success:
Kenyan Kipsigis
(Borgerhoff-Mulder, 2000)
82 men and 64 women in 4 income groups
No trade-off for men and for 3 of the 4 income groups
New Mexican men
(Kaplan et al., 1995)
Post-demographic transition
Empirics of Intertemporal Trade-off in Reproductive Success in Humans
Poorly identified
Unobserved heterogeneity
Reverse causality
Small samples
Often post-demographic transition populations
No intertemporal trade-off in reproductive success:
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 7 / 72
Kenyan Kipsigis
(Borgerhoff-Mulder, 2000)
82 men and 64 women in 4 income groups
No trade-off for men and for 3 of the 4 income groups
New Mexican men
(Kaplan et al., 1995)
Post-demographic transition
Empirics of Intertemporal Trade-off in Reproductive Success in Humans
Poorly identified
Unobserved heterogeneity
Reverse causality
Small samples
Often post-demographic transition populations
No intertemporal trade-off in reproductive success:
Kenyan Kipsigis
(Borgerhoff-Mulder, 2000)
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 7 / 72
82 men and 64 women in 4 income groups
No trade-off for men and for 3 of the 4 income groups
New Mexican men
(Kaplan et al., 1995)
Post-demographic transition
Empirics of Intertemporal Trade-off in Reproductive Success in Humans
Poorly identified
Unobserved heterogeneity
Reverse causality
Small samples
Often post-demographic transition populations
No intertemporal trade-off in reproductive success:
Kenyan Kipsigis
(Borgerhoff-Mulder, 2000)
82 men and 64 women in 4 income groups
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 7 / 72
No trade-off for men and for 3 of the 4 income groups
New Mexican men
(Kaplan et al., 1995)
Post-demographic transition
Empirics of Intertemporal Trade-off in Reproductive Success in Humans
Poorly identified
Unobserved heterogeneity
Reverse causality
Small samples
Often post-demographic transition populations
No intertemporal trade-off in reproductive success:
Kenyan Kipsigis
(Borgerhoff-Mulder, 2000)
82 men and 64 women in 4 income groups
No trade-off for men and for 3 of the 4 income groups
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 7 / 72
New Mexican men
(Kaplan et al., 1995)
Post-demographic transition
Empirics of Intertemporal Trade-off in Reproductive Success in Humans
Poorly identified
Unobserved heterogeneity
Reverse causality
Small samples
Often post-demographic transition populations
No intertemporal trade-off in reproductive success:
Kenyan Kipsigis
(Borgerhoff-Mulder, 2000)
82 men and 64 women in 4 income groups
No trade-off for men and for 3 of the 4 income groups
New Mexican men
(Kaplan et al., 1995)
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 7 / 72
Post-demographic transition
Empirics of Intertemporal Trade-off in Reproductive Success in Humans
Poorly identified
Unobserved heterogeneity
Reverse causality
Small samples
Often post-demographic transition populations
No intertemporal trade-off in reproductive success:
Kenyan Kipsigis
(Borgerhoff-Mulder, 2000)
82 men and 64 women in 4 income groups
No trade-off for men and for 3 of the 4 income groups
New Mexican men
(Kaplan et al., 1995)
Post-demographic transition
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 7 / 72
Historical Lab: Quebec During its Demographic Explosion
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 8 / 72
Quebec During its Demographic Explosion
Quebec 1608–1800
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 9 / 72
Founder population order of magnitude smaller than carrying capacity
Reproductive histories of 3,798 lineages
Genealogy of nearly 500,000 individuals
Quebec During its Demographic Explosion
Quebec 1608–1800
Founder population order of magnitude smaller than carrying capacity
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 9 / 72
Reproductive histories of 3,798 lineages
Genealogy of nearly 500,000 individuals
Quebec During its Demographic Explosion
Quebec 1608–1800
Founder population order of magnitude smaller than carrying capacity
Reproductive histories of 3,798 lineages
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 9 / 72
Genealogy of nearly 500,000 individuals
Quebec During its Demographic Explosion
Quebec 1608–1800
Founder population order of magnitude smaller than carrying capacity
Reproductive histories of 3,798 lineages
Genealogy of nearly 500,000 individuals
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 9 / 72
Findings
First comprehensive evidence for presence of an intertemporal trade-off in reproductive success (across any species)
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 10 / 72
An intermediate level of fecundity is associated with maximal long-run reproductive success.
Confirmation of selection of a quality strategy over the pre-industrial, pre-demographic transition era
The optimal level of fecundity is below the population mean
Findings
First comprehensive evidence for presence of an intertemporal trade-off in reproductive success (across any species)
An intermediate level of fecundity is associated with maximal long-run reproductive success.
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 10 / 72
Confirmation of selection of a quality strategy over the pre-industrial, pre-demographic transition era
The optimal level of fecundity is below the population mean
Findings
First comprehensive evidence for presence of an intertemporal trade-off in reproductive success (across any species)
An intermediate level of fecundity is associated with maximal long-run reproductive success.
Confirmation of selection of a quality strategy over the pre-industrial, pre-demographic transition era
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 10 / 72
The optimal level of fecundity is below the population mean
Findings
First comprehensive evidence for presence of an intertemporal trade-off in reproductive success (across any species)
An intermediate level of fecundity is associated with maximal long-run reproductive success.
Confirmation of selection of a quality strategy over the pre-industrial, pre-demographic transition era
The optimal level of fecundity is below the population mean
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 10 / 72
Main Hypothesis: Moderate Fertility Maximizes LR Reproductive Success
LR Reproductive Success
Galor and Klemp
Biocultural Origins of Human Capital
Children
July 4, 2015 11 / 72
Main Hypothesis: Moderate Fertility Maximizes LR Reproductive Success
LR Reproductive Success
Galor and Klemp
Biocultural Origins of Human Capital
Children
July 4, 2015 11 / 72
Empirical challenges: Unobserved Heterogeneity
Income and education
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 12 / 72
Transmitted intergenerationally
Positively affects fertility
Biases the results towards the absence of an intergenerational trade-off
Empirical challenges: Unobserved Heterogeneity
Income and education
Transmitted intergenerationally
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 12 / 72
Positively affects fertility
Biases the results towards the absence of an intergenerational trade-off
Empirical challenges: Unobserved Heterogeneity
Income and education
Transmitted intergenerationally
Positively affects fertility
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 12 / 72
Biases the results towards the absence of an intergenerational trade-off
Empirical challenges: Unobserved Heterogeneity
Income and education
Transmitted intergenerationally
Positively affects fertility
Biases the results towards the absence of an intergenerational trade-off
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 12 / 72
Challenge: Unobserved heterogeneity
Grandchildren
Galor and Klemp
Poor
Biocultural Origins of Human Capital
Children
July 4, 2015 13 / 72
Challenge: Unobserved heterogeneity
Grandchildren
Galor and Klemp
Ordinary
Children
Biocultural Origins of Human Capital
July 4, 2015 13 / 72
Challenge: Unobserved heterogeneity
Grandchildren
Galor and Klemp
Biocultural Origins of Human Capital
Rich
Children
July 4, 2015 13 / 72
Challenge: Unobserved heterogeneity
Grandchildren
Galor and Klemp
Biocultural Origins of Human Capital
Children
July 4, 2015 13 / 72
Empirical Challenge: Reverse causality
Effect of offspring quality on quantity
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 14 / 72
Lower quality of children that results in higher child mortality
Higher fertility via the replacement motive
Empirical Challenge: Reverse causality
Effect of offspring quality on quantity
Lower quality of children that results in higher child mortality
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 14 / 72
Higher fertility via the replacement motive
Empirical Challenge: Reverse causality
Effect of offspring quality on quantity
Lower quality of children that results in higher child mortality
Higher fertility via the replacement motive
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 14 / 72
Empirical strategy
Exploits random variation in fecundity (i.e., the physiological ability to reproduce) to establish the optimality of moderate fecundity for longrun reproductive success.
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 15 / 72
Employ the time interval between the date of marriage and the first birth
(TFB) as a source of variation in fecundity, .
TFB is affected by:
Genetic predisposition
Socio-environmental conditions
Random elements that affect conception
Identification will be based on the random variation in TFB
Empirical strategy
Exploits random variation in fecundity (i.e., the physiological ability to reproduce) to establish the optimality of moderate fecundity for longrun reproductive success.
Employ the time interval between the date of marriage and the first birth
(TFB) as a source of variation in fecundity, .
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 15 / 72
TFB is affected by:
Genetic predisposition
Socio-environmental conditions
Random elements that affect conception
Identification will be based on the random variation in TFB
Empirical strategy
Exploits random variation in fecundity (i.e., the physiological ability to reproduce) to establish the optimality of moderate fecundity for longrun reproductive success.
Employ the time interval between the date of marriage and the first birth
(TFB) as a source of variation in fecundity, .
TFB is affected by:
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 15 / 72
Genetic predisposition
Socio-environmental conditions
Random elements that affect conception
Identification will be based on the random variation in TFB
Empirical strategy
Exploits random variation in fecundity (i.e., the physiological ability to reproduce) to establish the optimality of moderate fecundity for longrun reproductive success.
Employ the time interval between the date of marriage and the first birth
(TFB) as a source of variation in fecundity, .
TFB is affected by:
Genetic predisposition
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 15 / 72
Socio-environmental conditions
Random elements that affect conception
Identification will be based on the random variation in TFB
Empirical strategy
Exploits random variation in fecundity (i.e., the physiological ability to reproduce) to establish the optimality of moderate fecundity for longrun reproductive success.
Employ the time interval between the date of marriage and the first birth
(TFB) as a source of variation in fecundity, .
TFB is affected by:
Genetic predisposition
Socio-environmental conditions
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 15 / 72
Random elements that affect conception
Identification will be based on the random variation in TFB
Empirical strategy
Exploits random variation in fecundity (i.e., the physiological ability to reproduce) to establish the optimality of moderate fecundity for longrun reproductive success.
Employ the time interval between the date of marriage and the first birth
(TFB) as a source of variation in fecundity, .
TFB is affected by:
Genetic predisposition
Socio-environmental conditions
Random elements that affect conception
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 15 / 72
Identification will be based on the random variation in TFB
Empirical strategy
Exploits random variation in fecundity (i.e., the physiological ability to reproduce) to establish the optimality of moderate fecundity for longrun reproductive success.
Employ the time interval between the date of marriage and the first birth
(TFB) as a source of variation in fecundity, .
TFB is affected by:
Genetic predisposition
Socio-environmental conditions
Random elements that affect conception
Identification will be based on the random variation in TFB
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 15 / 72
Introduction Empirical strategy
The Distribution of Time to First Birth (TFB)
5
4
3
2
1
0
0
Galor and Klemp
38
Weeks from marriage to first birth
Biocultural Origins of Human Capital
142
July 4, 2015 16 / 72
Introduction Empirical strategy
Time to First Birth (TFB) and the Number of Children
10
9.5
9
8.5
40
Galor and Klemp
Show unconditional relationship
60 80
Weeks from marriage to first birth
100
Biocultural Origins of Human Capital
120
July 4, 2015 17 / 72
Introduction Empirical strategy
Isolating the Random Variation in TFB
Control for socioeconomic factors
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 18 / 72
Literacy
Gender
Birth year, birth parish and death parish
Month of marriage and month of birth of firstborn
Marriage and stoppage age
Birth order
Lineage-specific fixed effects
Control for genetic predisposition
Lineage-specific fixed effects
Identification based on variation across siblings as opposed to the population as a whole
Introduction Empirical strategy
Isolating the Random Variation in TFB
Control for socioeconomic factors
Literacy
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 18 / 72
Gender
Birth year, birth parish and death parish
Month of marriage and month of birth of firstborn
Marriage and stoppage age
Birth order
Lineage-specific fixed effects
Control for genetic predisposition
Lineage-specific fixed effects
Identification based on variation across siblings as opposed to the population as a whole
Introduction Empirical strategy
Isolating the Random Variation in TFB
Control for socioeconomic factors
Literacy
Gender
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 18 / 72
Birth year, birth parish and death parish
Month of marriage and month of birth of firstborn
Marriage and stoppage age
Birth order
Lineage-specific fixed effects
Control for genetic predisposition
Lineage-specific fixed effects
Identification based on variation across siblings as opposed to the population as a whole
Introduction Empirical strategy
Isolating the Random Variation in TFB
Control for socioeconomic factors
Literacy
Gender
Birth year, birth parish and death parish
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 18 / 72
Month of marriage and month of birth of firstborn
Marriage and stoppage age
Birth order
Lineage-specific fixed effects
Control for genetic predisposition
Lineage-specific fixed effects
Identification based on variation across siblings as opposed to the population as a whole
Introduction Empirical strategy
Isolating the Random Variation in TFB
Control for socioeconomic factors
Literacy
Gender
Birth year, birth parish and death parish
Month of marriage and month of birth of firstborn
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 18 / 72
Marriage and stoppage age
Birth order
Lineage-specific fixed effects
Control for genetic predisposition
Lineage-specific fixed effects
Identification based on variation across siblings as opposed to the population as a whole
Introduction Empirical strategy
Isolating the Random Variation in TFB
Control for socioeconomic factors
Literacy
Gender
Birth year, birth parish and death parish
Month of marriage and month of birth of firstborn
Marriage and stoppage age
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 18 / 72
Birth order
Lineage-specific fixed effects
Control for genetic predisposition
Lineage-specific fixed effects
Identification based on variation across siblings as opposed to the population as a whole
Introduction Empirical strategy
Isolating the Random Variation in TFB
Control for socioeconomic factors
Literacy
Gender
Birth year, birth parish and death parish
Month of marriage and month of birth of firstborn
Marriage and stoppage age
Birth order
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 18 / 72
Lineage-specific fixed effects
Control for genetic predisposition
Lineage-specific fixed effects
Identification based on variation across siblings as opposed to the population as a whole
Introduction Empirical strategy
Isolating the Random Variation in TFB
Control for socioeconomic factors
Literacy
Gender
Birth year, birth parish and death parish
Month of marriage and month of birth of firstborn
Marriage and stoppage age
Birth order
Lineage-specific fixed effects
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 18 / 72
Control for genetic predisposition
Lineage-specific fixed effects
Identification based on variation across siblings as opposed to the population as a whole
Introduction Empirical strategy
Isolating the Random Variation in TFB
Control for socioeconomic factors
Literacy
Gender
Birth year, birth parish and death parish
Month of marriage and month of birth of firstborn
Marriage and stoppage age
Birth order
Lineage-specific fixed effects
Control for genetic predisposition
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 18 / 72
Lineage-specific fixed effects
Identification based on variation across siblings as opposed to the population as a whole
Introduction Empirical strategy
Isolating the Random Variation in TFB
Control for socioeconomic factors
Literacy
Gender
Birth year, birth parish and death parish
Month of marriage and month of birth of firstborn
Marriage and stoppage age
Birth order
Lineage-specific fixed effects
Control for genetic predisposition
Lineage-specific fixed effects
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 18 / 72
Identification based on variation across siblings as opposed to the population as a whole
Introduction Empirical strategy
Isolating the Random Variation in TFB
Control for socioeconomic factors
Literacy
Gender
Birth year, birth parish and death parish
Month of marriage and month of birth of firstborn
Marriage and stoppage age
Birth order
Lineage-specific fixed effects
Control for genetic predisposition
Lineage-specific fixed effects
Identification based on variation across siblings as opposed to the population as a whole
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 18 / 72
Introduction Empirical strategy
Reproductive Success of Head of Dynasties
Focus on head of dynasties born prior to 1685 allows for:
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 19 / 72
Reproductive histories of 3,798 head of dynasties
Tracing reproductive histories over 4 generations
Account for lineage-specific fixed effects
Introduction Empirical strategy
Reproductive Success of Head of Dynasties
Focus on head of dynasties born prior to 1685 allows for:
Reproductive histories of 3,798 head of dynasties
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 19 / 72
Tracing reproductive histories over 4 generations
Account for lineage-specific fixed effects
Introduction Empirical strategy
Reproductive Success of Head of Dynasties
Focus on head of dynasties born prior to 1685 allows for:
Reproductive histories of 3,798 head of dynasties
Tracing reproductive histories over 4 generations
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 19 / 72
Account for lineage-specific fixed effects
Introduction Empirical strategy
Reproductive Success of Head of Dynasties
Focus on head of dynasties born prior to 1685 allows for:
Reproductive histories of 3,798 head of dynasties
Tracing reproductive histories over 4 generations
Account for lineage-specific fixed effects
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 19 / 72
Head of Dynasty
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 20 / 72
Introduction Empirical strategy
The Lineage of Head of Dynasties
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 21 / 72
Introduction Empirical strategy
The Lineage of Head of Dynasties
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 22 / 72
Lineages of Head of Dynasties
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 23 / 72
Lineages of Head of Dynasties
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 23 / 72
Number of Children
10
8
6
4
2
0
1 5 10 15 20
Number of children
25 30
Number of Grandchildren
5
4
3
2
1
0
1 50 100
Number of grandchildren
150 200
Summary Statistics
Mean Median
(1) (2)
S.D.
(3)
Count
(4)
Children
Surviving children a
Grandchildren
Great-grandchildren
Great-great-grandchildren b
Years from marriage to first birth (TFB)
Literate
Fraction of literate children
Fraction of surviving children a
Fraction of surviving children with known literacy a
Age at first marriage
Age at last delivery
9.70
4.63
47.35
187.59
294.01
1.20
0.66
0.74
0.49
0.76
22.67
41.95
10
4
44
3.99
2.59
28.07
3,798
3,798
3,798
159 140.17
3,798
171 363.58
3,798
1.02
1
1
0.47
0.47
0.35
3,798
2,222
3,448
0.50
0.67
22.2
42.1
0.21
3,798
0.56
3,727
5.46
3,798
8.61
3,798 c
Survival is recorded at age 40.
b
The moderate increase in the mean and median number of descendants from the third to the fourth generation (i.e. from great-grandchildren to great-great-grandchildren) reflects the fact that these cohorts are less fully observed. Furthermore, since men produce children at later ages than women, this effect is more pronounced among men.
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 26 / 72
Children vs. TFB: restricted cubic splines
A
38
Weeks from marriage to first birth
142
Grandchildren vs. TFB: restricted cubic splines
B
38
Weeks from marriage to first birth
142
Great-Grandchildren vs. TFB: restricted cubic splines
C
38
Weeks from marriage to first birth
142
Great-Great-Grandchildren vs. TFB: restricted cubic splines
D
38
Weeks from marriage to first birth
142
Econometric Specification
The effect of TFB on the number of children: ln D i , 1
= β
0 , 1
+ β
1 , 1
TFB i
+ Z i
β
2 , 1
+ ε i , 1
,
The effect of TFB on the number of offspring in generation t = 2 , 3 , 4: ln D i , t
= β
0 , t
+ β
1 , t
TFB i
+ β
2 , t
TFB i
2
+ Z i
β
3 , t
+ ε i , t
,
Econometric Specification
The effect of TFB on the number of children: ln D i , 1
= β
0 , 1
+ β
1 , 1
TFB i
+ Z i
β
2 , 1
+ ε i , 1
,
The effect of TFB on the number of offspring in generation t = 2 , 3 , 4: ln D i , t
= β
0 , t
+ β
1 , t
TFB i
+ β
2 , t
TFB i
2
+ Z i
β
3 , t
+ ε i , t
,
TFB and Number of Descendants for Head of Lineages Born before 1685
(Accounting for Maternal Founder FE)
TFB
TFB 2
Literate
Male
Stoppage age fixed effects
Gen. 1
(1)
-.052**
(.024)
No
Gen. 2
(2)
.167
(.163)
-.068
(.053)
No
Gen. 3
(3)
Gen. 4
(4)
.505**
(.205)
.783***
(.264)
-.193*** -.310***
(.067) (.087)
No No
Log number of descendants in:
Gen. 1 Gen. 2 Gen. 3
(5)
-.062***
(.024)
-.006
(.030)
.220***
(.031)
No
(6)
.140
(.162)
-.063
(.053)
.063
(.040)
.254***
(.039)
No
(7)
.463**
(.204)
Gen. 4
(8)
.773***
(.264)
-.183*** -.309***
(.067) (.087)
.148***
(.051)
.138**
(.066)
.299***
(.047)
No
.131**
(.060)
No
Gen. 1
(9)
-.077***
(.011)
-.027*
(.014)
-.028*
(.015)
Yes
Number of observations
Adjusted R
2
Joint sign.-level of TFB & TFB 2
Maximizing TFB
Lower limit of 90% CI
Upper limit of 90% CI
3,798
.015
.031
3,798
.016
.196
1.224
-
-
3,798
.038
.002
1.307
.961
1.467
3,798
.306
.000
1.261
.999
1.398
3,798
.032
.010
3,798
.029
.130
1.113
-
-
3,798
.052
.002
1.263
.827
1.435
3,798
.307
.000
1.249
.976
1.389
3,798
.799
.000
This table presents the results of a series of fixed-effects regressions of the number of descendants in generation t on time to first birth measured in years, i.e.
TFB and TFB 2 . All regressions account for Maternal Founder fixed effects. Birth year and marriage age dummies are included as controls. Furthermore, stoppage age dummies are included in columns 9–10. A dummy indicating unknown literacy is included in the regressions underlying column 5–10. Standard errors clustered at the level of the firstborn are reported in parentheses. * p < 0.10, ** p < 0.05, *** p < 0.01.
Gen. 4
(10)
.810***
(.258)
-.325***
(.084)
.109*
(.066)
.036
(.063)
Yes
3,798
.355
.000
1.247
1.002
1.376
TFB and Number of Descendants for Head of Lineages Born 1685
1660–1685 (accounting for Maternal Founder FE)
TFB
TFB 2
Literate
Male
Stoppage age fixed effects
Gen. 1
(1)
-.065**
(.025)
No
Gen. 2
(2)
.237
(.175)
-.096*
(.057)
No
Gen. 3
(3)
.544**
(.216)
Gen. 4
Log number of descendants in:
Gen. 1 Gen. 2 Gen. 3
(4) (6) (7)
.830***
(.280)
-.214*** -.337***
(.071) (.092)
No No
(5)
-.075***
(.025)
-.031
(.033)
.219***
(.032)
No
.193
(.174)
-.086
(.057)
.023
(.042)
.279***
(.040)
No
.487**
(.216)
Gen. 4
(8)
.802***
(.282)
-.199*** -.330***
(.071) (.093)
.120**
(.054)
.116*
(.070)
.315***
(.049)
No
.177***
(.063)
No
Gen. 1
(9)
-.079***
(.012)
-.035**
(.015)
-.035**
(.015)
Yes
Number of observations
Adjusted R 2
Joint sign.-level of TFB & TFB 2
Maximizing TFB
Lower limit of 90% CI
Upper limit of 90% CI
3,376
.019
.010
3,376
.020
.066
1.230
-11.448
1.505
3,376
.042
.001
1.270
.923
1.427
3,376
.335
.000
1.232
.961
1.37
3,376
.036
.003
3,376
.036
.042
1.124
-
-
3,376
.058
.000
1.219
.761
1.392
3,376
.337
.000
1.214
.919
1.357
3,376
.804
.000
This table presents the results of a series of fixed-effects OLS regressions of the number of descendants in generation t on time to first birth measured in years, i.e.
TFB and TFB 2 for heads of lineages born in the period 1660–1685. All regressions account for Maternal Founder fixed effects. Birth year and marriage age dummies are included as controls. Furthermore, stoppage age dummies are included in columns 9–10. A dummy indicating unknown literacy is included in the regressions underlying column 5–10. Standard errors clustered at the level of the firstborn are reported in parentheses. * p < 0.10, ** p < 0.05, *** p < 0.01.
Gen. 4
(10)
.773***
(.277)
-.323***
(.091)
.087
(.070)
.115*
(.066)
Yes
3,376
.379
.000
1.196
.886
1.343
Robustness to Alternative Cohorts: No Restrictions on Cohorts (Accounting for Maternal Founder FE)
TFB
TFB 2
Literate
Male
Stoppage age fixed effects
Gen. 1
(1)
-.064***
(.017)
No
Gen. 2
(2)
.055
(.120)
-.039
(.040)
No
Gen. 3
(3)
.319**
(.141)
Gen. 4
Log number of descendants in:
Gen. 1 Gen. 2 Gen. 3
(4) (6) (7)
.409**
(.199)
-.133*** -.192***
(.046) (.065)
No No
(5)
-.071***
(.017)
.007
(.023)
.222***
(.018)
No
.031
(.119)
-.033
(.039)
.078***
(.030)
.220***
(.023)
No
.293**
(.140)
Gen. 4
(8)
.393**
(.199)
-.126*** -.187***
(.046) (.065)
.133***
(.037)
.134***
(.049)
.191***
(.027)
No
.088**
(.036)
No
Gen. 1
(9)
-.090***
(.008)
-.014
(.011)
-.013
(.009)
Yes
Number of observations
Adjusted R 2
Joint sign.-level of TFB & TFB 2
Maximizing TFB
Lower limit of 90% CI
Upper limit of 90% CI
7,664
.026
.000
7,664
.022
.015
.694
-
-
7,664
.067
.000
1.200
.741
1.376
7,664
.413
.000
1.065
.466
1.269
7,664
.049
.000
7,664
.036
.008
.457
-
-
7,664
.077
.000
1.166
.604
1.355
7,664
.414
.000
1.049
.396
1.261
7,664
.789
.000
This table presents the results of a series of fixed-effects OLS regressions of the number of descendants in generation t on time to first birth measured in years, i.e.
TFB and TFB 2 for heads of lineages born in the entire sample period. All regressions account for Maternal Founder fixed effects. Birth year and marriage age dummies are included as controls. Furthermore, stoppage age dummies are included in columns 9–10. A dummy indicating unknown literacy is included in the regressions underlying column 5–10. Standard errors clustered at the level of the firstborn are reported in parentheses. * p < 0.10, ** p < 0.05, *** p < 0.01.
Gen. 4
(10)
.459**
(.198)
-.212***
(.065)
.121**
(.048)
.019
(.038)
Yes
7,664
.431
.000
1.079
.610
1.264
Mechanisms
The observed patterns may reflect the positive effect of reduced fertility and thus higher child quality on the reproductive success of each child
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 35 / 72
Child quality enhances the likelihood that a child would
Reach the reproductive age and marry – a preconditions for reproductive success
Marry early – a preconditions for large number of children
Be educated and thus would have higher earning capacity and reproductive success
Mechanisms
The observed patterns may reflect the positive effect of reduced fertility and thus higher child quality on the reproductive success of each child
Child quality enhances the likelihood that a child would
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 35 / 72
Reach the reproductive age and marry – a preconditions for reproductive success
Marry early – a preconditions for large number of children
Be educated and thus would have higher earning capacity and reproductive success
Mechanisms
The observed patterns may reflect the positive effect of reduced fertility and thus higher child quality on the reproductive success of each child
Child quality enhances the likelihood that a child would
Reach the reproductive age and marry – a preconditions for reproductive success
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 35 / 72
Marry early – a preconditions for large number of children
Be educated and thus would have higher earning capacity and reproductive success
Mechanisms
The observed patterns may reflect the positive effect of reduced fertility and thus higher child quality on the reproductive success of each child
Child quality enhances the likelihood that a child would
Reach the reproductive age and marry – a preconditions for reproductive success
Marry early – a preconditions for large number of children
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 35 / 72
Be educated and thus would have higher earning capacity and reproductive success
Mechanisms
The observed patterns may reflect the positive effect of reduced fertility and thus higher child quality on the reproductive success of each child
Child quality enhances the likelihood that a child would
Reach the reproductive age and marry – a preconditions for reproductive success
Marry early – a preconditions for large number of children
Be educated and thus would have higher earning capacity and reproductive success
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 35 / 72
The Effect of TFB on the Fraction of Children Surviving to Age 40 and
Marry
(1)
Fraction of children surviving to age 40 that got married
(2) (3) (4)
TFB
Literate
Male
Stoppage age fixed effects
.299*** .256**
(.113) (.112)
.235**
(.111)
.232**
(.110)
.770*** .763*** .779***
(.110) (.110) (.109)
.396*** .365***
No No
(.112)
No
(.120)
Yes
Number of observations 3,727 3,727 3,727 3,727
This table presents the results of a series of fractional logit regressions of the fraction of children, i.e., individuals in the first generation, surviving to age 40 that got married on time to first birth measured in years, i.e., TFB , for heads of lineages with at least one child surviving to age 40. Birth year and marriage age dummies are included as controls. Furthermore, stoppage age dummies are included in column
4. A dummy indicating unknown literacy is included in the regressions underlying column 2–4. Standard errors clustered at the level of the firstborn are reported in parentheses. * p < 0.10, ** p < 0.05, *** p < 0.01.
The Effect of TFB on the on the Average Marriage Age
(1)
Average marriage age of children
(2) (3) (4)
TFB
Literate
Male
Stoppage age fixed effects
-.430*** -.400*** -.376**
(.005) (.008) (.013)
-.339**
(.023)
-.629*** -.621*** -.705***
(.001) (.001) (.000)
-.406*** -.720***
No No
(.009)
No
(.000)
Yes
Number of observations
Adjusted R
2
3,796
.006
3,796
.010
3,796
.011
3,796
.036
This table presents the results of a series of OLS regressions of the average marriage age of childre,, i.e., individuals in the first generation, on time to first birth measured in years, i.e., TFB . Birth year and marriage age dummies are included as controls.
Furthermore, stoppage age dummies are included in column 4. A dummy indicating unknown literacy is included in the regressions underlying column 2–4. Standard errors clustered at the level of the firstborn are reported in parentheses. * p < 0.10, ** p < 0.05,
*** p < 0.01.
The Effect of TFB on the on Share of Literate Offspring
(1)
Fraction of literate children
(2) (3) (4)
TFB
Literate
Male
Stoppage age fixed effects
.401*** .351***
(.090) (.090)
.322***
(.091)
.337***
(.091)
1.308*** 1.307*** 1.305***
No
(.094)
No
(.094)
.563***
(.090)
No
(.095)
.407***
(.098)
Yes
Number of observations 3,448 3,448 3,448 3,448
This table presents the results of a series of fractional logit regressions of the share of children, i.e., individuals in the first generation, obtaining literacy on time to first birth measured in years, i.e.,
TFB
, for heads of lineages with at least one surviving child with observed literacy status. Birth year and marriage age dummies are included as controls. Furthermore, stoppage age dummies are included in column 4. A dummy indicating unknown literacy is included in the regressions underlying column 2–4. Standard errors clustered at the level of the firstborn are reported in parentheses. *
p <
0.10, **
p <
0.05, ***
p <
0.01.
Findings
Maximal reproductive success is attained by couples with a moderate level of TFB
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 39 / 72
Lower TFB maximizes the number of children
An intermediate TFB maximizes long-run reproductive success
In comparison to highly fertile couples whose first child is born without any delays (i.e., 38 weeks after the marriage) they have:
0.3 fewer children
0.4 more grandchildren
8.4 more great-grandchildren
15.7 more great-great-grandchildren
Selection of a quality strategy
The optimal level of TFB and fertility is below the population median
Findings
Maximal reproductive success is attained by couples with a moderate level of TFB
Lower TFB maximizes the number of children
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 39 / 72
An intermediate TFB maximizes long-run reproductive success
In comparison to highly fertile couples whose first child is born without any delays (i.e., 38 weeks after the marriage) they have:
0.3 fewer children
0.4 more grandchildren
8.4 more great-grandchildren
15.7 more great-great-grandchildren
Selection of a quality strategy
The optimal level of TFB and fertility is below the population median
Findings
Maximal reproductive success is attained by couples with a moderate level of TFB
Lower TFB maximizes the number of children
An intermediate TFB maximizes long-run reproductive success
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 39 / 72
In comparison to highly fertile couples whose first child is born without any delays (i.e., 38 weeks after the marriage) they have:
0.3 fewer children
0.4 more grandchildren
8.4 more great-grandchildren
15.7 more great-great-grandchildren
Selection of a quality strategy
The optimal level of TFB and fertility is below the population median
Findings
Maximal reproductive success is attained by couples with a moderate level of TFB
Lower TFB maximizes the number of children
An intermediate TFB maximizes long-run reproductive success
In comparison to highly fertile couples whose first child is born without any delays (i.e., 38 weeks after the marriage) they have:
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 39 / 72
0.3 fewer children
0.4 more grandchildren
8.4 more great-grandchildren
15.7 more great-great-grandchildren
Selection of a quality strategy
The optimal level of TFB and fertility is below the population median
Findings
Maximal reproductive success is attained by couples with a moderate level of TFB
Lower TFB maximizes the number of children
An intermediate TFB maximizes long-run reproductive success
In comparison to highly fertile couples whose first child is born without any delays (i.e., 38 weeks after the marriage) they have:
0.3 fewer children
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 39 / 72
0.4 more grandchildren
8.4 more great-grandchildren
15.7 more great-great-grandchildren
Selection of a quality strategy
The optimal level of TFB and fertility is below the population median
Findings
Maximal reproductive success is attained by couples with a moderate level of TFB
Lower TFB maximizes the number of children
An intermediate TFB maximizes long-run reproductive success
In comparison to highly fertile couples whose first child is born without any delays (i.e., 38 weeks after the marriage) they have:
0.3 fewer children
0.4 more grandchildren
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 39 / 72
8.4 more great-grandchildren
15.7 more great-great-grandchildren
Selection of a quality strategy
The optimal level of TFB and fertility is below the population median
Findings
Maximal reproductive success is attained by couples with a moderate level of TFB
Lower TFB maximizes the number of children
An intermediate TFB maximizes long-run reproductive success
In comparison to highly fertile couples whose first child is born without any delays (i.e., 38 weeks after the marriage) they have:
0.3 fewer children
0.4 more grandchildren
8.4 more great-grandchildren
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 39 / 72
15.7 more great-great-grandchildren
Selection of a quality strategy
The optimal level of TFB and fertility is below the population median
Findings
Maximal reproductive success is attained by couples with a moderate level of TFB
Lower TFB maximizes the number of children
An intermediate TFB maximizes long-run reproductive success
In comparison to highly fertile couples whose first child is born without any delays (i.e., 38 weeks after the marriage) they have:
0.3 fewer children
0.4 more grandchildren
8.4 more great-grandchildren
15.7 more great-great-grandchildren
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 39 / 72
Selection of a quality strategy
The optimal level of TFB and fertility is below the population median
Findings
Maximal reproductive success is attained by couples with a moderate level of TFB
Lower TFB maximizes the number of children
An intermediate TFB maximizes long-run reproductive success
In comparison to highly fertile couples whose first child is born without any delays (i.e., 38 weeks after the marriage) they have:
0.3 fewer children
0.4 more grandchildren
8.4 more great-grandchildren
15.7 more great-great-grandchildren
Selection of a quality strategy
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 39 / 72
The optimal level of TFB and fertility is below the population median
Findings
Maximal reproductive success is attained by couples with a moderate level of TFB
Lower TFB maximizes the number of children
An intermediate TFB maximizes long-run reproductive success
In comparison to highly fertile couples whose first child is born without any delays (i.e., 38 weeks after the marriage) they have:
0.3 fewer children
0.4 more grandchildren
8.4 more great-grandchildren
15.7 more great-great-grandchildren
Selection of a quality strategy
The optimal level of TFB and fertility is below the population median
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 39 / 72
TFB and number of descendants for head of lineages born –1685
(accounting for Maternal Founder FE)
Gen. 1 Gen. 2 Gen. 3 Gen. 4 Gen. 1 Gen. 2 Gen. 3
Log number of descendants in:
Gen. 4 Gen. 1 Gen. 2 Gen. 3 Gen. 4 Gen. 1 Gen. 2 Gen. 3 Gen. 4
TFB
TFB
Literate
Male
2
Stoppage age fixed effects
(1)
-.052**
(.024)
No
(2)
.167
(.163)
-.068
(.053)
No
(3)
.505**
(.205)
-.193***
(.067)
No
(4)
.783***
(.264)
-.310***
(.087)
No
(5)
-.053**
(.024)
-.008
(.031)
No
(6)
.170
(.163)
-.070
(.053)
.060
(.040)
No
(7)
.499**
(.205)
-.191***
(.067)
.145***
(.051)
No
(8)
.788***
(.264)
-.313***
(.087)
.136**
(.066)
No
(9)
-.062***
(.024)
-.006
(.030)
.220***
(.031)
No
(10)
.140
(.162)
-.063
(.053)
.063
(.040)
.254***
(.039)
No
(11)
.463**
(.204)
-.183***
(.067)
.148***
(.051)
.299***
(.047)
No
(12)
.773***
(.264)
-.309***
(.087)
.138**
(.066)
.131**
(.060)
No
(13)
-.077***
(.011)
-.027*
(.014)
-.028*
(.015)
Yes
(14)
.208
(.130)
-.089**
(.042)
.044
(.032)
.025
(.031)
Yes
(15)
.535***
(.181)
-.210***
(.059)
.125***
(.046)
.085*
(.043)
Yes
Number of observations
Adjusted R
2
Joint sign.-level of TFB & TFB
2
Maximizing TFB
Lower limit of 90% CI
Upper limit of 90% CI
3,798
.015
.031
3,798
.016
.196
1.224
-
-
3,798
.038
.002
1.307
.961
1.467
3,798
.306
.000
1.261
.999
1.398
3,798
.017
.026
3,798
.016
.184
1.223
-
-
3,798
.041
.003
1.304
.948
1.466
3,798
.307
.000
1.260
1.000
1.397
3,798
.032
.010
3,798
.029
.130
1.113
-
-
3,798
.052
.002
1.263
.827
1.435
3,798
.307
.000
1.249
.976
1.389
3,798
.799
.000
3,798
.442
.002
1.163
-.141
1.393
3,798
.296
.000
1.272
1.012
1.403
3,798
.355
.000
1.247
1.002
1.376
This table presents the results of a series of fixed-effects regressions of the number of descendants in generation t on time to first birth measured in years, i.e.
TFB and TFB
2
. All regressions account for Maternal Founder fixed effects. Birth year and marriage age dummies are included as controls. Furthermore, stoppage age dummies are included in columns 13–16. A dummy indicating unknown literacy is included in the regressions underlying column 5–16. Standard errors clustered at the level of the firstborn are reported in parentheses. * p < 0.10, ** p < 0.05, *** p < 0.01.
(16)
.810***
(.258)
-.325***
(.084)
.109*
(.066)
.036
(.063)
Yes
TFB and number of descendants for head of lineages born 1660–1685
(accounting for Maternal Founder FE)
TFB
TFB
Male
2
Literate
Stoppage age fixed effects
Gen. 1
(1)
-.065**
(.025)
No
Gen. 2
(2)
.237
(.175)
-.096*
(.057)
No
Gen. 3
(3)
.544**
(.216)
-.214***
(.071)
No
Gen. 4
(4)
.830***
(.280)
-.337***
(.092)
No
Gen. 1
(5)
-.066***
(.025)
-.028
(.033)
No
Gen. 2
(6)
.238
(.175)
-.097*
(.057)
.027
(.043)
No
Gen. 3
Log number of descendants in:
Gen. 4 Gen. 1 Gen. 2
(7)
.537**
(.217)
-.212***
(.071)
.124**
(.055)
No
(8)
.830***
(.282)
-.337***
(.093)
.118*
(.070)
No
(9)
-.075***
(.025)
-.031
(.033)
.219***
(.032)
No
(10)
.193
(.174)
-.086
(.057)
.023
(.042)
.279***
(.040)
No
Gen. 3
(11)
.487**
(.216)
-.199***
(.071)
.120**
(.054)
.315***
(.049)
No
Gen. 4
(12)
.802***
(.282)
-.330***
(.093)
.116*
(.070)
.177***
(.063)
No
Gen. 1
(13)
-.079***
(.012)
-.035**
(.015)
-.035**
(.015)
Yes
Gen. 2
(14)
.201
(.139)
-.088*
(.046)
.016
(.035)
.056*
(.032)
Yes
Gen. 3
(15)
.498***
(.191)
-.203***
(.063)
.110**
(.049)
.116***
(.044)
Yes
Gen. 4
(16)
.773***
(.277)
-.323***
(.091)
.087
(.070)
.115*
(.066)
Yes
Number of observations
Adjusted R
2
Joint sign.-level of TFB & TFB
2
Maximizing TFB
Lower limit of 90% CI
Upper limit of 90% CI
3,376
.019
.010
3,376
.020
.066
1.230
-11.448
1.505
3,376
.042
.001
1.270
.923
1.427
3,376
.335
.000
1.232
.961
1.37
3,376
.020
.009
3,376
.020
.066
1.230
-11.038
1.504
3,376
.045
.001
1.268
.910
1.427
3,376
.335
.000
1.231
.959
1.369
3,376
.036
.003
3,376
.036
.042
1.124
-
-
3,376
.058
.000
1.219
.761
1.392
3,376
.337
.000
1.214
.919
1.357
3,376
.804
.000
3,376
.451
.005
1.137
-1.02
1.387
3,376
.306
.000
1.224
.883
1.375
This table presents the results of a series of fixed-effects OLS regressions of the number of descendants in generation t on time to first birth measured in years, i.e.
TFB and TFB
2 for heads of lineages born in the period 1660–1685. All regressions account for
Maternal Founder fixed effects. Birth year and marriage age dummies are included as controls. Furthermore, stoppage age dummies are included in columns 13–16. A dummy indicating unknown literacy is included in the regressions underlying column 5–16.
Standard errors clustered at the level of the firstborn are reported in parentheses. * p < 0.10, ** p < 0.05, *** p < 0.01.
3,376
.379
.000
1.196
.886
1.343
Robustness to alternative cohorts: no restrictions on cohorts (accounting for
Maternal Founder FE)
TFB
TFB
Male
2
Literate
Stoppage age fixed effects
Gen. 1
(1)
-.064***
(.017)
No
Gen. 2
(2)
.055
(.120)
-.039
(.040)
No
Gen. 3
(3)
.319**
(.141)
-.133***
(.046)
No
Gen. 4
(4)
.409**
(.199)
-.192***
(.065)
No
Gen. 1
(5)
-.064***
(.017)
.007
(.023)
No
Gen. 2
(6)
.050
(.120)
-.038
(.040)
.078**
(.030)
No
Gen. 3
Log number of descendants in:
Gen. 4 Gen. 1 Gen. 2
(7)
.310**
(.141)
-.129***
(.046)
.133***
(.037)
No
(8)
.400**
(.199)
-.189***
(.065)
.133***
(.049)
No
(9)
-.071***
(.017)
.007
(.023)
.222***
(.018)
No
(10)
.031
(.119)
-.033
(.039)
.078***
(.030)
.220***
(.023)
No
Gen. 3
(11)
.293**
(.140)
-.126***
(.046)
.133***
(.037)
.191***
(.027)
No
Gen. 4
(12)
.393**
(.199)
-.187***
(.065)
.134***
(.049)
.088**
(.036)
No
Gen. 1
(13)
-.090***
(.008)
-.014
(.011)
-.013
(.009)
Yes
Gen. 2
(14)
.096
(.094)
-.060*
(.031)
.057**
(.024)
.005
(.019)
Yes
Gen. 3
(15)
.378***
(.125)
-.158***
(.041)
.110***
(.033)
.019
(.025)
Yes
Gen. 4
(16)
.459**
(.198)
-.212***
(.065)
.121**
(.048)
.019
(.038)
Yes
Number of observations
Adjusted R
2
Joint sign.-level of TFB & TFB
2
Maximizing TFB
Lower limit of 90% CI
Upper limit of 90% CI
7,664
.026
.000
7,664
.022
.015
.694
-
-
7,664
.067
.000
1.200
.741
1.376
7,664
.413
.000
1.065
.466
1.269
7,664
.026
.000
7,664
.023
.018
.663
-
-
7,664
.070
.000
1.198
.708
1.378
7,664
.414
.000
1.060
.434
1.268
7,664
.049
.000
7,664
.036
.008
.457
-
-
7,664
.077
.000
1.166
.604
1.355
7,664
.414
.000
1.049
.396
1.261
7,664
.789
.000
7,664
.420
.000
.798
-3.176
1.167
7,664
.282
.000
1.196
.924
1.333
7,664
.431
.000
1.079
.610
1.264
This table presents the results of a series of fixed-effects OLS regressions of the number of descendants in generation t on time to first birth measured in years, i.e.
TFB and TFB
2 for heads of lineages born in the entire sample period. All regressions account for Maternal Founder fixed effects. Birth year and marriage age dummies are included as controls. Furthermore, stoppage age dummies are included in columns 13–16. A dummy indicating unknown literacy is included in the regressions underlying column
5–16. Standard errors clustered at the level of the firstborn are reported in parentheses. * p < 0.10, ** p < 0.05, *** p < 0.01.
Summary statistics for females
(1) (2)
Mean Median
(3)
S.D.
(4)
Count
Children
Grandchildren
Great-grandchildren
Great-great-grandchildren
Years from marriage to first birth (TFB)
Literate
Fraction of literate children
Fraction of surviving children b
Fraction of surviving children with known literacy b
Age at first marriage
Age at last delivery
9.42
45.99
187.65
341.04
1.23
0.68
0.72
0.59
0.62
19.34
38.27
10
43
3.66
27.40
2,058
2,058
159 142.74
2,058
206.5
408.07
2,058
1.04
1
1
0.60
0.60
18.7
40.3
0.49
2,058
0.47
1,192
0.36
1,872
0.20
2,058
0.40
2,044
3.79
2,058
6.46
2,058 a
The moderate increase in the mean and median number of descendants from the third to the fourth generation (i.e. from great-grandchildren to great-great-grandchildren) reflects the fact that these cohorts are less fully observed. Furthermore, since men produce children at lager ages than women, this effect is more pronounced among men.
b
Survival is recorded at the average marriage age, i.e. 23 years.
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 44 / 72
Summary statistics for males
(1) (2)
Mean Median
(3)
S.D.
(4)
Count
Children
Grandchildren
Great-grandchildren
Great-great-grandchildren
Years from marriage to first birth (TFB)
Literate
Fraction of literate children
Fraction of surviving children b
Fraction of surviving children with known literacy b
Age at first marriage
Age at last delivery
10.03
48.94
187.53
238.38
1.16
0.64
0.76
0.60
0.57
26.62
46.31
10
45
4.32
28.77
1,740
1,740
159 137.10
1,740
136.5
293.17
1,740
0.99
1
1
0.60
0.50
25.9
46.9
0.44
1,740
0.48
1,030
0.34
1,576
0.20
1,740
0.39
1,728
4.41
1,740
8.81
1,740 a
The moderate increase in the mean and median number of descendants from the third to the fourth generation (i.e. from great-grandchildren to great-great-grandchildren) reflects the fact that these cohorts are less fully observed. Furthermore, since men produce children at lager ages than women, this effect is more pronounced among men.
b
Survival is recorded at the average marriage age, i.e. 23 years.
Galor and Klemp
Biocultural Origins of Human Capital
July 4, 2015 45 / 72
TFB and number of descendants for head of lineages born –1685 (using
GLM w/ neg. binom.)
Gen. 1 Gen. 2 Gen. 3 Gen. 4 Gen. 1
Number of descendants in:
Gen. 2 Gen. 3 Gen. 4 Gen. 1 Gen. 2 Gen. 3 Gen. 4 Gen. 1 Gen. 2 Gen. 3 Gen. 4
TFB
TFB
Literate
Male
2
Stoppage age fixed effects
(1)
-.063***
(.018)
No
(2)
.120
(.148)
-.047
(.048)
No
(3)
.371**
(.185)
-.133**
(.060)
No
(4)
.616***
(.230)
-.232***
(.074)
No
(5)
-.063***
(.018)
-.002
(.020)
No
(6)
.090
(.148)
-.039
(.049)
.143***
(.030)
No
(7)
.324*
(.186)
-.119*
(.061)
.222***
(.038)
No
(8)
.573**
(.231)
-.219***
(.074)
.227***
(.046)
No
(9)
-.081***
(.017)
-.009
(.019)
.295***
(.019)
No
(10)
.028
(.145)
-.026
(.048)
.138***
(.030)
.379***
(.027)
No
(11)
.260
(.184)
-.106*
(.060)
.219***
(.038)
.402***
(.035)
No
(12)
.525**
(.231)
-.209***
(.074)
.225***
(.046)
.306***
(.041)
No
(13)
-.090***
(.010)
-.056***
(.011)
-.004
(.011)
Yes
(14)
.049
(.121)
-.035
(.039)
.098***
(.025)
.094***
(.023)
Yes
(15)
.346**
(.167)
-.137**
(.054)
.180***
(.035)
.150***
(.032)
Yes
Number of observations
Joint sign.-level of TFB & TFB
2
Maximizing TFB
3,798
.000
3,798
.446
1.273
3,798
.059
1.399
3,798
.000
1.328
3,798
.000
3,798
.445
1.159
3,798
.086
1.363
3,798
.001
1.306
3,798
.000
3,798
.139
.546
3,798
.041
1.229
3,798
.000
1.255
3,798
.000
3,798
.016
.702
3,798
.001
1.262
3,798
.000
1.277
This table presents the results of a series of GLM regressions, with a negative binomial distribution and a logarithmic link function, of the number of descendants in generation t on time to first birth measured in years, i.e.
TFB and TFB
2
. Birth year and marriage age dummies are included as controls. Furthermore, stoppage age dummies are included in columns 13–16. A dummy indicating unknown literacy is included in the regressions underlying column 5–16. Standard errors clustered at the level of the firstborn are reported in parentheses. * p < 0.10, ** p < 0.05, *** p < 0.01.
(16)
.573**
(.227)
-.224***
(.073)
.195***
(.047)
.194***
(.042)
Yes
TFB and number of descendants for head of lineages born –1685 (using
GLM w/ neg. binom. – accounting for Maternal Founder FE)
Gen. 1 Gen. 2 Gen. 3 Gen. 4 Gen. 1
Number of descendants in:
Gen. 2 Gen. 3 Gen. 4 Gen. 1 Gen. 2 Gen. 3 Gen. 4 Gen. 1 Gen. 2 Gen. 3 Gen. 4
TFB
TFB
Literate
Male
2
Stoppage age fixed effects
(1)
-.041**
(.020)
No
(2)
.126
(.146)
-.054
(.047)
No
(3)
.412**
(.179)
-.162***
(.058)
No
(4)
.547**
(.215)
-.234***
(.068)
No
(5)
-.042**
(.020)
-.006
(.026)
No
(6)
.131
(.146)
-.056
(.047)
.059*
(.036)
No
(7)
.414**
(.178)
-.163***
(.058)
.121***
(.045)
No
(8)
.557***
(.215)
-.238***
(.069)
.119**
(.056)
No
(9)
-.052***
(.020)
-.006
(.025)
.248***
(.027)
No
(10)
.101
(.145)
-.050
(.047)
.061*
(.036)
.280***
(.036)
No
(11)
.387**
(.177)
-.158***
(.058)
.124***
(.045)
.282***
(.044)
No
(12)
.547**
(.215)
-.236***
(.068)
.121**
(.056)
.127**
(.054)
No
(13)
-.073***
(.010)
-.020
(.013)
-.015
(.014)
Yes
(14)
.195*
(.118)
-.087**
(.038)
.055*
(.029)
.044
(.029)
Yes
(15)
.468***
(.160)
-.191***
(.052)
.118***
(.040)
.083**
(.040)
Yes
Number of observations
Joint sign.-level of TFB & TFB
2
Maximizing TFB
3,798
.045
3,798
.207
1.172
3,798
.001
1.271
3,798
.000
1.169
3,798
.037
3,798
.189
1.174
3,798
.001
1.27
3,798
.000
1.169
3,798
.009
3,798
.097
1.01
3,798
.000
1.226
3,798
.000
1.155
3,798
.000
3,798
.000
1.116
3,798
.000
1.225
3,798
.000
1.159
This table presents the results of a series of GLM regressions, with a negative binomial distribution and a logarithmic link function, of the number of descendants in generation t on time to first birth measured in years, i.e.
TFB and TFB
2
. All regressions include dummies for Maternal Founder fixed effects (the results without the Maternal Founder Fixed Effects is presented in Table A.7). Birth year and marriage age dummies are included as controls. Furthermore, stoppage age dummies are included in columns 13–16. A dummy indicating unknown literacy is included in the regressions underlying column 5–16. Standard errors clustered at the level of the firstborn are reported in parentheses. * p < 0.10, ** p < 0.05, *** p < 0.01.
(16)
.595***
(.216)
-.256***
(.069)
.104*
(.055)
.072
(.055)
Yes
Robustness to including extinct lineages – accounting for Maternal Founder fixed effects
Log number of descendants in:
Gen. 1 Gen. 2 Gen. 3 Gen. 4
Log 1+number of descendants in:
Gen. 1 Gen. 2 Gen. 3 Gen. 4
(1) (2) (3) (4) (5) (6) (7) (8)
TFB
TFB
2
Literate
Male
-.046*
(.026)
.151
(.172)
.462**
(.208)
.773***
(.264)
-.071
(.057)
.060
-.180*** -.309***
(.068)
.153***
(.087)
.138** -.022
(.033) (.042) (.053)
.321*** .321*** .349***
(.033) (.040) (.050)
(.066)
.131**
(.060)
-.041*
(.021)
.414*
(.215)
.822*** 1.169***
(.294) (.336)
-.149** -.293*** -.427***
(.068)
.058
(.093)
.144*
(.108)
.180** -.014
(.027) (.056) (.076)
.276*** .456*** .597***
(.027) (.051) (.069)
(.087)
.478***
(.076)
Number of observations
Adjusted R
2
Joint sign.-level of TFB & TFB 2
Maximizing TFB
4,240
.052
.073
4,002
.044
.083
1.067
3,933
.068
.003
1.283
3,798
.307
.000
1.249
4,240
.054
.053
4,240
.065
.046
1.387
4,240
.084
.002
1.403
4,240
.241
.000
1.368
This table presents the results of a series of fixed-effects regressions of the number of descendants in generation t on time to first birth measured in years, i.e.
TFB and TFB
2
. All regressions account for Maternal Founder fixed effects. Birth year and marriage age dummies are included as controls. A dummy indicating unknown literacy is also included in the regressions. In columns 1–4, the restriction of at least one observed great-great-grandchild is relaxed and extinct lineages drop out of the sample in the relevant generations. In columns 5–8, the same extended sample is used, but the outcome is ln(1 + D i , t
), where the added number 1 ensures that the logarithmic transformation is defined and all lineages remains included. Standard errors clustered at the level of the firstborn are reported in parentheses. * p < 0.10, ** p < 0.05, *** p < 0.01.
Robustness to additional sample restrictions: more than one child and marriage occurring after turning 15 years – accounting for Maternal
Founder fixed effects
Gen. 1
(1)
Gen. 2
(2)
Log number of descendants in:
Gen. 3 Gen. 4 Gen. 1 Gen. 2
(3) (4) (5) (6)
Gen. 3
(7)
Gen. 4
(8)
TFB
TFB 2
Literate
Male
-.053**
(.021)
.204
(.157)
-.082
(.051)
.560***
(.201)
-.211*** -.344***
(.066) (.086)
.025
(.027)
.090**
(.038)
.175***
(.050)
.251*** .272*** .314***
(.029) (.038) (.047)
.891*** -.052**
(.263) (.025)
.162**
(.065)
.135**
(.061)
-.016
(.031)
.129
(.175)
-.057
(.058)
.052
(.041)
.461**
(.218)
.139***
(.052)
.225*** .269*** .318***
(.031) (.039) (.048)
.779***
(.279)
-.180** -.310***
(.073) (.093)
.123*
(.068)
.153**
(.061)
Number of observations
Adjusted R 2
Joint sign.-level of TFB & TFB
2
Maximizing TFB
Lower limit of 90% CI
Upper limit of 90% CI
3,738
.044
.011
3,738
.037
.104
1.248
-
-
3,738
.060
.001
1.323
1.059
1.464
3,738
.318
.000
1.296
1.095
1.417
3,604
.035
.038
3,604
.031
.294
1.140
-
-
3,604
.055
.009
1.282
.798
1.458
3,604
.314
.000
1.256
.973
1.396
This table presents the results of a series of fixed-effects regressions of the number of descendants in generation t on time to first birth measured in years, i.e.
TFB and TFB
2
. All regressions account for Maternal Founder fixed effects. Birth year and marriage age dummies are included as controls.
A dummy indicating unknown literacy is also included in the regressions. In columns 1–4, the sample is restricted to heads of lineages who produced at least two children. In columns 5–8, the sample is restricted to heads of lineages who were married after their 15th birthday. Standard errors clustered at the level of the firstborn are reported in parentheses. * p < 0.10, ** p < 0.05, *** p < 0.01.
Robustness to additional control variables: number of marriages and spousal migration (accounting for Maternal Founder FE)
Gen. 1
(1)
Gen. 2
(2)
Log number of descendants in:
Gen. 3 Gen. 4 Gen. 1 Gen. 2
(3) (4) (5) (6)
Gen. 3
(7)
Gen. 4
(8)
TFB
TFB
Male
2
Literate
Total number of marriages fixed effects
Total number of marriages of spouse fixed effects
Immigration status of spouse fixed effects
Emigration status of spouse fixed effects
-.059***
(.022)
-.013
(.028)
.168***
(.028)
Yes
Yes
No
No
.282*
(.157)
.590***
(.200)
.828***
(.263)
-.109** -.225*** -.328***
(.051) (.066) (.087)
.059
(.039)
.143***
(.050)
.135**
(.066)
.204*** .255***
(.037) (.046)
Yes
Yes
Yes
Yes
No
No
No
No
.116*
(.060)
Yes
Yes
No
No
-.059**
(.024)
.117
(.161)
.423**
(.204)
.720***
(.264)
-.007
(.030)
-.054
(.052)
.067*
(.040)
-.168**
(.067)
.148***
(.051)
.198*** .206*** .232***
(.032) (.039) (.047)
No
No
No
No
No
No
Yes
Yes
Yes
Yes
Yes
Yes
-.289***
(.087)
.140**
(.066)
.037
(.061)
No
No
Yes
Yes
Number of observations
Adjusted R
2
Joint sign.-level of TFB & TFB
2
Maximizing TFB
Lower limit of 90% CI
Upper limit of 90% CI
3,798
.188
.009
3,798
.116
.027
1.289
.420
1.509
3,798
.097
.000
1.311
1.063
1.446
3,798
.312
.000
1.263
1.024
1.393
3,798
.035
.013
3,798
.035
.169
1.073
-
-
3,798
.062
.005
1.260
.716
1.445
3,798
.315
.000
1.244
.935
1.392
This table presents the results of a series of fixed-effects regressions of the number of descendants in generation t on time to first birth measured in years, i.e.
TFB and
TFB 2 . All regressions account for Maternal Founder fixed effects. Birth year and marriage age dummies are included as controls. A dummy indicating unknown literacy is also included in the regressions. In columns 1–4, dummies for the total number of marriages experienced during the lifetime of the heads of lineages, as well as dummies for the total number of marriages experienced by the first spouses of the heads of lineages, are included. In columns 5–8, dummies indicating the immigration and emigration statuses of the head of the first spouses of the heads of lineages are included. Standard errors clustered at the level of the firstborn are reported in parentheses. * p < 0.10,
** p < 0.05, *** p < 0.01.
Robustness to gender distinction – sample restricted to females (accounting for Maternal Founder FE)
Gen. 1
(1)
Gen. 2
(2)
Gen. 3
Log number of descendants in:
Gen. 4 Gen. 1 Gen. 2
(3) (4) (5) (6)
Gen. 3
(7)
Gen. 4
(8)
TFB
TFB
2
Literate
-.079***
(.030)
.374*
(.214)
.985***
(.278)
-.140** -.347***
(.069) (.088)
1.568*** -.080***
(.349) (.030)
-.560***
(.109)
-.040
(.043)
.380*
(.215)
.993***
(.278)
-.142** -.349***
(.069) (.088)
.072
.120
(.067) (.084)
1.581***
(.349)
-.564***
(.110)
.145
(.107)
Number of observations
Adjusted R
2
Joint sign.-level of TFB & TFB
2
Maximizing TFB
Lower limit of 90% CI
Upper limit of 90% CI
2,058
.089
.008
2,058
.064
.057
1.334
.366
1.58
2,058
.091
.000
1.418
1.24
1.538
2,058
.267
.000
1.401
1.263
1.500
2,058
.089
.007
2,058
.064
.056
1.339
.428
1.583
2,058
.092
.000
1.422
1.247
1.542
2,058
.267
.000
1.402
1.267
1.501
This table presents the results of a series of fixed-effects regressions of the number of descendants in generation t on time to first birth measured in years, i.e.
TFB and TFB
2 for female heads of lineages. All regressions account for Maternal Founder fixed effects. Birth year and marriage age dummies are included as controls. A dummy indicating unknown literacy is also included in the regressions. Standard errors clustered at the level of the firstborn are reported in parentheses. * p < 0.10, ** p < 0.05, *** p < 0.01.
Robustness to additional control variables: location of birth and death – accounting for Maternal Founder fixed effects
Gen. 1 Gen. 2 Gen. 3
Log number of descendants in:
Gen. 4 Gen. 1 Gen. 2 Gen. 3 Gen. 4
TFB
TFB
Male
2
Literate
Constant
Birth parish fixed effects
Death parish fixed effects
(1)
-0.062
∗∗∗
(0.024)
-0.008
(0.031)
0.214
∗∗∗
(0.031)
3.861
∗∗∗
(0.366)
Yes
No
(2)
0.106
(0.163)
-0.052
(0.053)
0.062
(0.040)
0.253
∗∗∗
(0.039)
5.949
∗∗∗
(0.649)
Yes
No
(3)
0.416
∗∗
(0.205)
-0.168
∗∗
(0.067)
0.149
∗∗∗
(0.051)
0.298
∗∗∗
(0.048)
6.288
∗∗∗
(0.667)
Yes
No
(4)
0.695
∗∗∗
(0.267)
-0.284
∗∗∗
(0.088)
0.144
∗∗
(0.066)
0.134
∗∗
(0.061)
6.228
∗∗∗
(0.935)
Yes
No
(5)
-0.062
∗∗∗
(0.023)
-0.015
(0.031)
0.203
∗∗∗
(0.031)
2.256
∗∗∗
(0.428)
No
Yes
(6)
0.032
(0.161)
-0.023
(0.053)
0.037
(0.040)
0.208
∗∗∗
(0.038)
4.130
∗∗∗
(0.681)
No
Yes
(7)
0.391
∗
(0.202)
-0.154
∗∗
(0.066)
0.097
∗
(0.052)
0.226
∗∗∗
(0.046)
5.229
∗∗∗
(0.703)
No
Yes
Number of observations
Adjusted R
2
Joint sign.-level of TFB & TFB
2
Maximizing TFB
Lower limit of 90% CI
Upper limit of 90% CI
3,798
0.034
.009
3,798
0.030
.155
1.024
-
-
3,798
0.050
.002
1.234
.646
1.424
3,798
0.308
.000
1.223
.875
1.379
3,798
0.066
.009
3,798
0.098
.425
.700
-
-
3,798
0.130
.011
1.264
.607
1.461
This table presents the results of a series of fixed-effects regressions of the number of descendants in generation t on time to first birth measured in years, i.e.
TFB and TFB 2 . All regressions account for Maternal Founder fixed effects. Birth year and marriage age dummies are included as controls. A dummy indicating unknown literacy is also included in the regressions. In columns 1–4, dummies indicating the birth (or baptism) parish of the heads of lineages are included. In columns 5–8 dummies indicating the death (or burial) parish of the heads of lineages are included. Standard errors clustered at the level of the firstborn are reported in parentheses. * p < 0.10, ** p < 0.05, *** p < 0.01.
3,798
0.370
.000
1.243
.892
1.402
(8)
0.653
∗∗
(0.255)
-0.262
∗∗∗
(0.084)
0.074
(0.067)
0.033
(0.059)
4.676
∗∗∗
(0.921)
No
Yes
Robustness to additional control variables: month of marriage and first birth
– accounting for Maternal Founder fixed effects
TFB
TFB
Male
2
Literate
Month of marriage fixed effects
Month of birth of firstborn fixed effects
Gen. 1
(1)
-0.061
∗∗
(0.024)
-0.008
(0.030)
0.224
∗∗∗
(0.031)
Yes
No
Gen. 2
(2)
0.151
(0.163)
-0.067
(0.053)
0.062
(0.040)
0.255
∗∗∗
(0.039)
Yes
No
Gen. 3
Log number of descendants in:
Gen. 4 Gen. 1 Gen. 2
(3)
0.482
∗∗
(0.205)
-0.190
∗∗∗
(0.067)
0.146
∗∗∗
(0.051)
0.296
∗∗∗
(0.048)
Yes
No
(4)
0.798
∗∗∗
(0.265)
-0.318
∗∗∗
(0.087)
0.137
∗∗
(0.067)
0.123
∗∗
(0.060)
Yes
No
(5)
-0.050
∗∗
(0.024)
-0.004
(0.030)
0.220
(0.031)
No
Yes
∗∗∗
(6)
0.181
(0.166)
-0.073
(0.054)
0.065
(0.040)
0.257
∗∗∗
(0.039)
No
Yes
Gen. 3
(7)
0.500
∗∗
(0.208)
-0.194
∗∗∗
(0.068)
0.153
∗∗∗
(0.051)
0.298
∗∗∗
(0.047)
No
Yes
Number of observations
Adjusted R
2
Joint sign.-level of TFB & TFB
2
Maximizing TFB
Lower limit of 90% CI
Upper limit of 90% CI
3,798
0.032
.010
3,798
0.028
.109
1.126
-
-
3,798
0.051
.001
1.267
.865
1.433
3,798
0.307
.000
1.253
.994
1.389
3,798
0.032
.037
3,798
0.032
.169
1.242
-
-
3,798
0.056
.001
1.286
.904
1.452
This table presents the results of a series of fixed-effects regressions of the number of descendants in generation t on time to first birth measured in years, i.e.
TFB and TFB 2 . All regressions account for Maternal Founder fixed effects. Birth year and marriage age dummies are included as controls. A dummy indicating unknown literacy is also included in the regressions. In columns 1–4, dummies indicating the months of marriage of the heads of lineages are included. In columns 5–8, dummies indicating the month of birth of the the firstborn of the heads of lineages are included. Standard errors clustered at the level of the firstborn are reported in parentheses. * p < 0.10, ** p < 0.05, *** p < 0.01.
3,798
0.308
.000
1.254
.968
1.398
Gen. 4
(8)
0.787
∗∗∗
(0.274)
-0.314
∗∗∗
(0.089)
0.140
∗∗
(0.067)
0.129
∗∗
(0.060)
No
Yes
Robustness to additional control variable: birth order (accounting for
Maternal Founder FE)
Gen. 1
(1)
Gen. 2
(2)
Gen. 3
Log number of descendants in:
Gen. 4 Gen. 1 Gen. 2
(3) (4) (5) (6)
Gen. 3
(7)
Gen. 4
(8)
TFB
TFB
Literate
Male
2
Firstborn
Birth order fixed effects
-.062***
(.024)
-.006
(.030)
.219***
(.031)
.016
(.023)
No
.139
(.162)
-.062
(.053)
.064
(.040)
.250***
(.039)
.064**
(.031)
No
.462**
(.204)
.773***
(.264)
-.183*** -.309***
(.067) (.087)
.149*** .137**
(.051)
.296***
(.066)
.132**
(.047)
.042
(.038)
No
(.060)
-.019
(.047)
No
-.062***
(.024)
-.008
(.030)
.220***
(.031)
Yes
.144
(.162)
-.063
(.053)
.064
(.040)
.248***
(.039)
Yes
.456**
(.204)
.149***
(.051)
.293***
(.047)
Yes
.752***
(.265)
-.181*** -.303***
(.067) (.087)
.141**
(.067)
.131**
(.061)
Yes
Number of observations
Adjusted R 2
Joint sign.-level of TFB & TFB 2
Maximizing TFB
Lower limit of 90% CI
Upper limit of 90% CI
3,798
.032
.010
3,798
.030
.135
1.117
-
-
3,798
.052
.002
1.264
.828
1.436
3,798
.307
.000
1.249
.976
1.388
3,798
.031
.009
3,798
.030
.139
1.133
-
-
3,798
.052
.002
1.261
.807
1.434
3,798
.308
.000
1.240
.949
1.384
This table presents the results of a series of fixed-effects regressions of the number of descendants in generation t on time to first birth measured in years, i.e.
TFB and TFB 2 . All regressions account for Maternal Founder fixed effects. Birth year and marriage age dummies are included as controls.
A dummy indicating unknown literacy is also included in the regressions. In columns 1–4, a dummy for the firstborn status of the heads of lineages is included. In columns 5–8, dummies the birth order of the heads of lineages are included. Standard errors clustered at the level of the firstborn are reported in parentheses. * p < 0.10, ** p < 0.05, *** p < 0.01.
Effect of number of children on long-run reproductive success – accounting for Maternal Founder fixed effects
Gen. 2
(1)
Gen. 3
(2)
Gen. 4
(3)
Log number of descendants in:
Gen. 2 Gen. 3 Gen. 4
(4) (5) (6)
Gen. 2
(7)
Gen. 3
(8)
Gen. 4
(9)
Log Number of Children
(Log Number of Children)
Literate
Male
2
2.334***
(.524)
-.371
(.229)
3.446***
(.842)
-.728**
(.367)
4.491***
(1.078)
-1.280***
(.474)
2.339***
(.525)
-.376
(.230)
.068
(.047)
3.459***
(.851)
-.738**
(.373)
.151**
(.070)
4.545***
(1.091)
-1.306***
(.481)
.171*
(.092)
2.384***
(.520)
-.388*
(.228)
.067
(.047)
.129***
(.046)
3.515***
(.837)
-.749**
(.366)
.150**
(.070)
.266***
(.071)
4.619***
(1.075)
-1.324***
(.474)
.170*
(.092)
.307***
(.092)
Number of observations
First stage F (Kleibergen-Paap)
Joint sign. of linear and squared terms
Maximizing number of log(1+D)
Exp(Maximizing log(1+D))
Lower limit of 90% CI
Upper limit of 90% CI
4,240
30.701
.000
3.142
22.15
1.618
6.012
4,240
30.701
.000
2.367
9.665
1.770
8.556
4,240
30.701
.000
1.754
4.777
1.473
2.805
4,240
33.094
.000
3.113
21.488
1.473
2.805
4,240
33.094
.000
2.343
9.412
1.756
8.445
4,240
33.094
.000
1.74
4.697
1.465
2.757
4,240
31.891
.000
3.069
20.52
2.081
57.776
4,240
31.891
.000
2.345
9.433
1.767
7.475
4,240
31.891
.000
1.744
4.72
1.472
2.702
This table presents the results of a series of fixed-effects 2SLS regressions of the log number of descendants in generation t , i.e. ln(1 + D that the head of household i has in generations t , t = 2 , 3 , i , t
4, on the number of children and the number of children squared, i.e., ln(1 +
), where D
D i , 1 i , t is the number of descendants
) and (ln(1 + D i , 1
))
2
, instrumented by variation in TFB . The added number 1 ensures that the logarithmic transformation is defined for extinct lineages and all that lineages remain in the sample. All regressions account for Maternal Founder fixed effects. Birth year, marriage age, and stoppage age dummies are included as controls. Since the second stage of these 2SLS regressions is quadratic in the endogenous regressor, it is necessary to instrument for both the linear and the squared terms in order to identify the parameters. Thus, following Wooldridge (2010), pp. 267–268, a zeroth stage is introduced to the analysis, where D i , 1
1 , i is first regressed on TFB
1 , i and all the second-stage controls to obtain predicted values of the number of children. The predicted
) as well as (ln(1 + ˆ
1 , i
))
2
, and these transformed terms are then used as excluded instruments in the second stage. A dummy indicating unknown literacy is included in the regressions underlying columns 4–9. Standard errors clustered at the level of the firstborn are reported in parentheses. * p < 0.10, ** p < 0.05, *** p < 0.01.
15
10
5
0
0
Galor and Klemp
Age
Biocultural Origins of Human Capital
41 60
July 4, 2015 56 / 72
2.4
2.35
2.3
A
2.25
2.2
38 142
Weeks from marriage to first birth
5.1
5
C
4.9
4.8
4.7
38 142
Weeks from marriage to first birth
3.8
3.75
B
3.7
3.65
3.6
38 142
Weeks from marriage to first birth
5.4
5.2
5
D
4.8
4.6
38 142
Weeks from marriage to first birth
10
8
6
4
2
0
1 5 10
Number of children
15 20
10
8
6
4
2
0
1 5 10 15
Number of children
20 25 30
5
4
3
2
1
0
1 50 100
Number of grandchildren
150 200
5
4
3
2
1
0
1 50 100
Number of grandchildren
150 200
20
10
40
30
A
0
1 200 400 600 800
Number of great-grandchildren
1000
60 D
40
20
0
1 1000 2000 3000
Number of great-great-grandchildren
4000
40
20
20
10
40
30
B
0
1 200 400 600 800
Number of great-grandchildren
1000
60 E
0
1 1000 2000 3000
Number of great-great-grandchildren
4000
40
20
20
10
40
30
0
1
C
200 400 600 800
Number of great-grandchildren
1000
60 F
0
1 1000 2000 3000
Number of great-great-grandchildren
4000
Age at first marriage (All)
8
6
4
2
0
10 15 20 25 30 35 40 45
Age at marriage (years)
50 55 60 65 70
Age at first marriage (Females)
10
8
6
4
2
0
10 15 20 25 30 35 40 45
Age at marriage (years)
50 55 60 65 70
Age at first marriage (Males)
6
4
2
10
8
0
10 15 20 25 30 35 40 45 50
Age at marriage (years)
55 60 65 70
15
10
5
0
0
Age
41 60
10
9.5
9
8.5
40
60 80
Weeks from marriage to first birth
100 120
2.20
2.15
2.10
38 60 80 100
Weeks from marriage to first birth
4.92
4.90
4.88
4.86
4.84
38 60 80 100
Weeks from marriage to first birth
3.65
3.64
3.63
3.62
3.61
38 60 80 100
Weeks from marriage to first birth
4.95
4.90
4.85
4.80
38 60 80 100
Weeks from marriage to first birth
2.20
2.15
2.10
2.05
38 60 80 100
Weeks from marriage to first birth
4.95
4.90
4.85
4.80
38 60 80 100
Weeks from marriage to first birth
3.66
3.64
3.62
3.60
38 60 80 100
Weeks from marriage to first birth
4.95
4.90
4.85
4.80
4.75
38 60 80 100
Weeks from marriage to first birth
2.20
2.00
38 60 80 100 120
Weeks from marriage to first birth
5.00
4.90
4.80
4.70
38 60 80 100 120
Weeks from marriage to first birth
3.70
3.65
3.60
3.55
38 60 80 100 120
Weeks from marriage to first birth
5.00
4.90
4.80
4.70
4.60
38 60 80 100 120
Weeks from marriage to first birth
2.30
2.20
2.10
2.00
38 60 80 100 120
Weeks from marriage to first birth
5.00
4.90
4.80
4.70
38 60 80 100 120
Weeks from marriage to first birth
3.70
3.65
3.60
3.55
38 60 80 100 120
Weeks from marriage to first birth
5.20
5.00
4.80
4.60
38 60 80 100 120
Weeks from marriage to first birth
.3
.2
.1
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
Number of children
