Skip to main content

Journal of Population Sciences

Table 4 Checks on the relation between fertility and excess female survival at various old ages

From: Fertility decline and the emergence of excess female survival in post-reproductive ages in Italy

 

Dependent variable

ln (EFS75–79)

ln (EFS85–89)

ln (EFS95–99)

Estimate

SE

p value

Estimate

SE

p value

Estimate

SE

p value

All eight cohorts (1862–1866, …, 1932–1936)

 Intercept

0.774

0.104

< 0.001

1.072

0.114

< 0.001

1.713

0.157

< 0.001

 Fertility (If)

− 1.112

0.142

< 0.001

− 1.865

0.256

< 0.001

− 3.336

0.399

< 0.001

 Survival

− 0.409

0.118

< 0.001

− 0.104

0.205

0.612

7.734

2.157

< 0.001

 

df = 93

R2 = 0.46

S = 0.43

df = 93

R2 = 0.54

S = 0.67

df = 77

R2 = 0.70

S = 0.93

Three youngest cohorts (1912–1916, 1922–1926, 1932–1936)

 Intercept

0.727

0.109

< 0.001

0.698

0.193

0.001

0.861

0.254

0.006

 Fertility (If)

-0.511

0.124

< 0.001

-0.669

0.282

0.013

-1.389

0.394

0.002

 Survival

− 0.685

0.106

< 0.001

− 0.667

0.300

0.035

4.281

4.158

0.325

 Male LCM

0.003

0.001

0.001

0.007

0.002

0.002

0.011

0.003

0.004

 Female LCM

− 0.006

0.003

0.019

− 0.009

0.010

0.192

− 0.008

0.022

0.361

 

df = 43

R2 = 0.78

S = 0.09

df = 27

R2 = 0.71

S = 0.18

df = 11

R2 = 0.84

S = 0.53

  1. Regression of EFSx (excess female survival) at various ages (75–79 years; 85–89 years; 95–99 years). The main independent variable is Princeton’s index of general fertility (If), used as a proxy for cohort fertility, while survival is introduced as a control. The upper part of the table refers to all the cohorts in our dataset. The lower part considers only the three youngest cohorts, for which we can add an extra control on lung cancer mortality (LCM) at 75–79 years, used as a proxy for the prevalence of smoking. Standard errors (SE) have been computed with the heteroskedasticity- and autocorrelation-consistent sandwich estimator (Cribari-Neto, 2004). The p values for the coefficients of “If”, “male lung mortality” and “female lung mortality” refer to a one-tail t test. For each regression, we show the degrees of freedom (df), the explained variance (R2) and Shapiro’s test on the normality of regression residuals (S), where S < 0.05 signals a probable departure from normality