Care Provision through lifetime

1. Let’s begin by simulating data

This is a panel dataset that conatins 1000 observations and two periods.

Time independent variables:

  • Indiv: is the identification of each observation.
  • Gender: 1 is male and 0 female. For this study, the non-binary were already excluded.
  • Socioeconomic level: this is dummy variable that divides the population into three categories.

Time dependent variables:

  • Age: will be measured in the two periods. It should be noted that in the first we have observations of people that were between 18 and 35. The second round of data was taken 20 years after.
  • Labor: this is a binary variable that captures labor force participation.
  • Care: this is a binary variable that captures if the person provides informal care more than 10 hours a week.
  • Wages: level of labor income for those that are in the labor force.
#Lets randomly generate all the variable I want
indiv<-1:1000
gender<-sample(x = 0:1, size = 1000, replace = TRUE)
age_1<-sample(x = 18:35, size = 1000, replace = TRUE)
soceco_level<-sample(x = 1:3, size = 1000, replace = TRUE)
labor_1<-sample(x = 0:1, size = 1000, replace = TRUE)
care_1<-sample(x = 0:1, size = 1000, replace = TRUE)
labor_2<-sample(x = 0:1, size = 1000, replace = TRUE)
care_pref<-sample(x = 0:1, size = 1000, replace = TRUE)
shock_wage_1<-rnorm(1000, -5, 5)
shock_wage_2<-rnorm(1000, -5, 5)

#Format is as a data frame
my_data<-matrix(c(indiv, gender, age_1, soceco_level,labor_1,care_1,labor_2, care_pref, shock_wage_1, shock_wage_2 ), ncol=10)
my_data02<-as.data.frame.matrix(my_data)

#Rename the variables
my_data02<- 
  rename(my_data02,
    indiv = V1,
    gender = V2,
    age_1 = V3,
    soceco_level = V4,
    labor_1 = V5,
    care_1 = V6,
    labor_2 = V7,
    care_pref = V8,
    shock_wage_1 =V9,
    shock_wage_2 =V10
    )

my_data02<-mutate(my_data02,
                wage_1 = labor_1*exp(0.5*gender + 0.5*soceco_level + 0.05*age_1 - 0.025*age_1^2  + shock_wage_1),
                age_2 = age_1 + 20)

my_data02<-mutate(my_data02,
                wage_2 = labor_2*(wage_1 + exp(labor_1 + 0.05*age_2 - 0.025*age_2^2 + shock_wage_2)),
                care_2 = ifelse(care_1 - labor_1*((wage_1 - mean(wage_1))/mean(wage_1)) - gender + care_pref > 1, 1, 0))

table01<-(head(my_data02))
kable(table01, align = "rrrrrrrrrrr", digits = c(1, 1, 2, 1, 1, 4, 4,6,2,6,1))
indiv gender age_1 soceco_level labor_1 care_1 labor_2 care_pref shock_wage_1 shock_wage_2 wage_1 age_2 wage_2 care_2
1 1 32 3 1 0 0 1 -10.09 -10.630265 0 52 0 0
2 0 22 1 1 1 0 1 -18.34 -5.163540 0 42 0 1
3 1 20 2 1 0 1 1 2.48 -10.655038 0 40 0 0
4 1 28 1 1 1 1 1 2.38 -3.072351 0 48 0 1
5 1 33 2 0 1 0 1 -4.94 -3.441887 0 53 0 0
6 1 28 3 0 0 1 0 0.09 0.407149 0 48 0 0 
#xtable(my_data02[1:10, ])

2. Data observation

In this section I provide a table that computes the proportion of labor force participation and care provision by age.

#Lets randomly generate all the variable I want
table02<-my_data02 %>%
          group_by(age_1) %>%
            summarize( 
              lab_1 = mean(labor_1),
              car_1 = mean(care_1),
              lab_2 = mean(labor_2),
              car_2 = mean(care_2))

kable(table02, align = "rrrrrrr", digits = c(2, 1, 1, 5, 1, 1, 5))
age_1 lab_1 car_1 lab_2 car_2
18 0.5 0.6 0.56364 0.3
19 0.5 0.4 0.44828 0.3
20 0.5 0.5 0.44262 0.2
21 0.5 0.5 0.53226 0.4
22 0.5 0.6 0.52941 0.3
23 0.5 0.6 0.52381 0.4
24 0.5 0.6 0.40816 0.4
25 0.5 0.4 0.54688 0.3
26 0.5 0.5 0.48276 0.3
27 0.5 0.5 0.50980 0.4
28 0.6 0.6 0.46809 0.3
29 0.5 0.4 0.59677 0.3
30 0.5 0.6 0.46512 0.4
31 0.5 0.3 0.56140 0.3
32 0.6 0.6 0.55319 0.4
33 0.4 0.5 0.61290 0.3
34 0.6 0.4 0.34848 0.2
35 0.5 0.5 0.65909 0.3 
print(xtable(table02, type = "latex"), file = "table02.tex")

3. Does taking care when young affect taking care when adult?

Let’s run the following regression:

\[care_2 = \beta_0 + \beta_1 care_1 + \beta_2 X + \varepsilon\] Where the set of controls \(X\) contains: age, gender, wage_1, labor_1.

model01 <-lm(care_2 ~ care_1 + age_2 + gender+ wage_1 + labor_1, data=my_data02)
tab_model(model01,  p.style = "stars")
  care_2
Predictors Estimates CI
(Intercept) 0.17 -0.01 – 0.36
care_1 0.38 *** 0.34 – 0.42
age_2 -0.00 -0.01 – 0.00
gender -0.38 *** -0.43 – -0.34
wage_1 -0.14 ** -0.25 – -0.04
labor_1 0.39 *** 0.35 – 0.43
Observations 1000
R2 / R2 adjusted 0.502 / 0.500
  • p<0.05   ** p<0.01   *** p<0.001 
stargazer(model01, title="Results", align=TRUE, out = "table03.tex")
% Table created by stargazer v.5.2.2 by Marek Hlavac, Harvard University. E-mail: hlavac at fas.harvard.edu % Date and time: Thu, Mar 25, 2021 - 7:19:42 PM % Requires LaTeX packages: dcolumn