Economic Relationships in Data - Analysis of School Finances and Local Economy

Subset Data to Keep only 2015 Years

For this section we will analyze data from 2015. First, we will merge additional economic data to the data we used in problem set 1, only keeping 2015 data. Then we will explore the relationship between school finances and the local economy.

Q1a. Load districtdata.csv in your workstation and name the data frame districts. Subset the data to keep only 2015 years using the subset function, for example:

Districts2015 <— subset (districts, districts$year == 2015).

Next, create two variables from problem set 2: per-pupil spending and property tax share.

Q1b. Load the supplementary data for PS3, districtdata_ps3.cav and statedata_ps3.csv. Name the data frames data2015 and states2015, respectively.

Q1c. Merge the data. First, merge the state data to the 2015 district data. Create a tem-porary data frame, ie temp2015 4— merge(districts2015,states2015, by = "statecode")

Q1d. Merge temp2015 to data2015. Name the new data frame 0.1112015.

Q1e. Let's look at the state income distribution. First use the descr function and report the summary statistics for the state income distribution variable: stateinc2015. (Hint: only use the states2015 dataframe). Then, plot the density distribution of state income using the following commands:

Stateinc 4— density(states2015$stateinc2015) plot (stateinc)

Q1f. Repeat 2e, this time using fu112015 dataframe and districtinc2015. Include screenshots of both the state and district income distributions. Compare the summary statistics of the two income measures and both density distributions. Discuss the skew (compare mean ver-sus median), and any anomalies or outliers you see in the plots.

Q1g. The homeownership rate is a statistic defined as the fraction of occupied housing that is owned instead of rented. Create the district homeownership rate variable in the fu112015 data frame using the variables ownhousing2015 and occhousing2015

Q1h. Create three scatter plots to explore the relationship between district homeownership and other economic variables: per-pupil spending, taxshare, and district income. The com-mand for scatter plots is:

 
Plot (ful120155homerate, fu1120155pp_spend2015).

Make sure you use the variable names you created. Repeat this command for the scatter plots between homeownership and taxshare, then homeownership and district income. In-clude all three screenshots. Make sure homeownership rate is on the x-axis.

Q1i. Visually the relationship can be hard to discern with a lot of data points. Now, compute the correlation statistic of homeownership rate and the three variables in 2i. The command is: cor(fu112015Shomerate, fu112015$pp_spend2015,).

Q1j. The article discussion is around a qualitative belief that homeownership will lead to better public goods and civic participation. For example, some believe that neighborhoods with higher homeownership have better funded schools simply because people owning homes means they are more invested in the neighborhood. Discuss your findings in 2i and 2j. What role does income play in this conversation? Although we do not measure it here, what about the role of wealth?

Create three variables using the "fu112015" dataframe:

1. state unemployment rate: state unemployed in 2015 divided by state labor force in 2015.

2. fraction of population 25+ with a bachelors degree or highers: statebachplus25p2015 divided by statepop25p2015

3. State revenue per-pupil: Total revenues - total taxes, then divided by enrollment. Produce three scatter plots: 1. state median income (stateinc2015) vs state revenue per-pupil)