Because of the high economic growth rate and the general prosperity of the population in the 20th century in the U.S., many foreigners, who wanted to live the American Dream, were attracted to migrate to the U.S.
These droves up to the high number of migrants for the U.S. resulted in fewer job opportunities. All the controversy over immigration reform in recent years have sparked our team interest in the subject, so we sought out data to analyze in search of the linkages between the U.S. unemployment rates and immigration rates.
Our problem: To analyze the correlations between immigration growth and unemployment rate growth for different group of people, including native, blacks/African Americans, whites, high educated U.S. citizens, and low educated U.S. citizens.
We are eeking the correlations between immigration growth and unemployment rate growth for different group of people, including native, blacks/African Americans, whites, high educated U.S. citizens, and low educated U.S. citizens.
In order to determine state-by-state changes in the immigration rate of the U.S., we obtained immigration population data in the years 2000 and 2010 from IPUMS-USA.
Because we cannot access IPUMS, which required login, directly, we export these data to Excel, and uploaded them to personal space for sharing to the public, and use them for our project.
Then, we calculated the percentage change of immigration between year 2000 and 2010. Similarly, we calculated percentage change of unemployment rate year 2000 and 2010 for all different groups. Then, we use these percentage changes data to analyze the correlation between immigration and different group of data to determine whether immigration affect to U.S. unemployment rate.
Due to the difficulty of getting the raw data from the school department, we have changed our topic to this one. We spent too much time one the previous topic, but everything turned out to useless. Finally, we turned out just have little time to do this new question.
We chose 2000 and 2010 as our project year because this is the most recent decade data that we can get from the IPUMS. In our project, there are five major variables that will be compared to the change of immigration, and each of these variables also has multiple sub-variables. They are showed as follows.
1. States: 50 states in United States.
2. Native Unemployment rate
3. White American unemployment rate
4. African American Unemployment rate
5. High educated people’s unemployment rate
6. Low educated people’s unemployment rate
We combine all the variables together to conduct the correlation between changes of immigration to change of different group of people’s unemployment rate.
For visualization, we included bar chart and scatter plots to clearly show the unemployment rate in each variables. At the end, we included a big picture to summarise all the graph together to show the relationship.
from IPython.display import display
from IPython.display import Image
c = Image(filename='./visualizations/Correlation.png')
aa = Image(filename='./visualizations/African American.png')
a = Image(filename='./visualizations/Changes.png')
n = Image(filename='./visualizations/Native.png')
h = Image(filename='./visualizations/US Citizens with High Education Level.png')
l = Image(filename='./visualizations/US Citizens with Low Education Level.png')
w = Image(filename='./visualizations/White.png')
Correlations between each variables with immigration grouth.
Including:
* Correlation between Immigration Growth and the Unemployment Rate growth of native
* Correlation between Immigration Growth and the Unemployment Rate Growth of Afican American
* Correlation between Immigration Growth and the Unemployment Rate Growth of White
* Correlation between Immigration Growth and the Unemployment Rate Growth of US Citizens with High Education Level
* Correlation between Immigration Growth and the Unemployment Rate Growth of US Citizens with Low Education Level
display(c)
From the bar chart above, the correlation between percentage point change of native unemployment rate and percentage point change of immigration was around 0.30, for white natives was 0.35, for black or African American native was around 0.09, for high educated native was around 0.36, and low educated native was around 0.347. Since all these correlation values are close to 0 and far from 1.
There have five graphs, each of their x-axis represents percentage point change of immigration between 2000 and 2010 and the y-axis represents different groups of percentage point change for unemployment rate between 2000 and 2010. The first of which is of a non-grouped (total) native unemployed population and the following graphs each represent a grouped native unemployed population. The last two figures represents percentage point change of Mexican immigration to percentage point change of low educated native unemployment and percentage point change of Black native unemployment rate.
display(n)
Plot1:
This plot shows the relation between percentage point change of immigration to percentage point change of native unemployment rate between year 2000 and 2010.
The data points in the plot we do not look to show a trend and has an R^2 value of around 0.09; the calculated correlation value of around 0.3 backs up the visual assessment. These two values are far from 1, an indication of a perfect relationship.
But these values show that there is very little correlation between changes of immigration percentage to changes of native unemployment rate in year 2000 to 2010. Immigration did not strongly affect the native unemployment rate during the last decade.
display(w)
Plot2:
This plot is much like the plot1, but now only displays the native unemployment rate point percentage change in the y-axis.
From the graph we notice that R^2 for the two variables is around 0.12, and correlation between two variables is around 0.346. These numbers indicate that two variables have a low relationship like figure 1 because both plot1 and plot2 have similar R^2 and correlation values.
From our analysis the unemployment rate for white natives is not affected by immigration.
display(aa)
Plot3:
Plot3 displays the correlation between percentage point change of immigration to percentage point change of unemployment rate for native black or African American.
R^2 for these two variables is around 0.0088 and correlation is around 0.094. From these numbers we can conclude that these variables are almost 0, which indicates no correlation at all with each other.
But unlike previous plots, native African American population has much lower correlation and R^2 values. We noticed that immigration has almost no affect on Black or African American native unemployment rate. These two variables are totally independent of each other.
display(h)
Plot4:
In this plot, change of immigration percentage and change of high-educated unemployment rate for each state is spread randomly.
R^2 is around 0.132 and correlation is around 0.363. These numbers show that both variables have a weak relationship.
From the R^2 and the correlation value we see that the change of immigration has no affect on the unemployment rate for high-educated natives.
display(l)
Plot5:
The plot shows the change of immigration percentage versus the change of unemployment rate for the low educated natives. Most states randomly spotted and the dots are not close enough to form a regression line.
For these two variables R^2 is around 0.120 and correlation is around 0.347.
These numbers emphasize that unemployment of low educated natives are not affected by changes in immigration.
From five scatter plots, we are easy to find that there have no particular relationship between each two elements. It's means for every States, every different native citizes or education level persons has different changes when the number of immigration groth. This type of cloud is characheristic of normally distributed independent variable. So there are no specific relationship between the rate growth of every kind of citizens(inclue white, African American, people who have a hight education, and persons who have a low education) and immigration growth.
display(a)
After examining all the unemployment data for each state as a whole and different social classes of people like white native, black or African American native, high educated natives and low educated natives, we conclude that there is no relation between immigration versus unemployment for natives.
All the correlation and R2 values tell us that changes in immigration percentage from last decade are independent to change in unemployment of general natives and different group of natives.
All five scatter plots show us, the relationship between the rate growth of every kind of citizens(inclue white, African American, people who have a hight education, and persons who have a low education) and immigration growth are independent.
List the team members contributing to this notebook, along with their responsabilities: