Introduction

During the 20th century, the United States’ underwent multiple societal changes. From multiple civil rights movements for minorities to the civil rights acts for gender equality. During this century of social change; societies environments have altered from women independence to work to the makeup of a family unit. The data that was collected is a longitudinal analysis of the divorce rates between the years 1920 to 1996.

The short analysis, was conducted to examine the relationship between wife’s economic independence and divorce. I hypothesize that year and females entering the work force will heighten the rate of divorce. I also hypothesize that as births increase there will be a decline in divorce as the growth of a family increases unions and togetherness.

The variables that I have chosen to present my analysis are as follows

Independent Variables:

  • Femlab- Percent female participation in labor force aged 16+
  • Year- The year from 1920-1996
  • Birth- Births per 1000 women aged 15-44.

Dependent Variable

  • Divorce- divorce per 1000 women aged 15 or more.

Regression Models and Best Fit

d1<- lm(divorce~femlab, data= divusa)
d2<- lm(divorce~femlab + year, data= divusa)
d3<- lm(divorce~ femlab+ birth+ year, data= divusa)
d4<- lm(divorce~ femlab*birth, data= divusa)

**Comparison of Models

Dependent variable:
divorce
(1) (2) (3) (4)
femlab 0.439*** 0.753*** 0.594*** 0.784***
(0.023) (0.134) (0.120) (0.117)
birth -0.074*** 0.121**
(0.015) (0.056)
year -0.167** -0.122*
(0.070) (0.062)
femlab:birth -0.005***
(0.001)
Constant -3.655*** 312.119** 236.721** -9.914**
(0.928) (132.759) (116.242) (4.782)
Observations 77 77 77 77
R2 0.829 0.841 0.882 0.895
Adjusted R2 0.827 0.837 0.877 0.891
Residual Std. Error 2.361 (df = 75) 2.291 (df = 74) 1.989 (df = 73) 1.873 (df = 73)
F Statistic 363.142*** (df = 1; 75) 195.676*** (df = 2; 74) 181.440*** (df = 3; 73) 207.862*** (df = 3; 73)
Note: p<0.1; p<0.05; p<0.01

Model 1: Femlab Effect
In this model, the effect of females in the workforce on the rate of divorce were examined. The results shows that for every 1 year females enter in the workforce the rate of divorce increases by .438. This is statistically significant but, yet again I believe that there are other factors which interact with the year that have affected this to be true. These results are statistically significant at p <2e-16.

Model 2: Female in workforce and Year Effects
Here, we examined the effect of females in the workforce and the year on the average rate of divorce. The results show that as the unit of females entering the workforce increases the rate of divorce increases to .7526 while controlling the units of year divorce decreases by .1675. These results are not statistically significant.

Model 3: Femlab, Birth, and Year Effects
Here, we examined the effect of females in the workforce, Births of children, and year on the rate of divorce. Birth is shown to be a statistically significant p < 0.00000358 factor that influences the rate of divorce; This result is irrespective of year, which remains a statistically insignificant factor on the rate of divorce. Also, femlab remains a statistically significant influence.

Model 4: Interaction of Females in the workforce and childrens birth
Here the interaction between females in the workforce and births of children on the rate of divorce was examined. This interaction shows a effect of the interaction of p < 0.000394.

*Model 4 is the best fit model for analysis of the interaction of females in the workforce and birth of children on divorce rates, as it has the lowest AIC value.

Visualizations

These plots illustrate Model 4, the results of which showed that the interaction between females in the workforce and births of children are statistically significant influences of the rate of divorce.

Scatterplot

library(lemon)
ggplot(divusa, aes(x=year, y=divorce)) + 
  geom_point() + 
  coord_capped_cart(bottom='both', left='none') +
  theme_light() + theme(panel.border=element_blank(), axis.line = element_line())

Scatterplot w/ Regression Line

Interactive Plot

ggplotly(divorce_plot)
## We recommend that you use the dev version of ggplot2 with `ggplotly()`
## Install it with: `devtools::install_github('hadley/ggplot2')`

##Conclusion

From my short analysis and visualization’s, women entering the workforce and the number of children a woman bears were found to be the the most significant factors that explain the rate of divorce through the years. The higher the percentage of females working , the rate of divorce increases but, when a child is added to the family unit this rate decreases. From the interactive scatter plot, we have also examined that women are bearing less children as the years have gone by. This scatter plot suggests that there may be other motivations to the rate of divorce.