PEOPLE BOXES, INC.

APARTMENT COMPLEX CONSORTIUM PLANNING EXPANSION BASED ON DEMOGRAPHIC CHANGES

BRIEF

People Boxes, Inc. is a consortium of real estate owners who seek out growing real estate demand among young, affluent, highly-educated workers entering the workforce or relocating to new cities. They specialize in condominium-style new construction, saving on costs by reusing blueprints in different cities and working with national contracting firms. Their revenue growth depends on identifying new markets to expand into, filling a niche in high-demand, high-income cities.

They have collected data on demographics and income in a number of metropolitan statistical areas (MSAs) in the United States and would like assistance in analyzing the data to provide some background information and some conclusions on the underlying relationships of income and age, education levels, and rental prevalence, as well as the determinants and effects of the supply of housing stock.

1. Youth

- Are America's "young" cities poorer than the median US income? For this, define "young" as "between 14 and 24", and find the proportion of each MSA that is young. Then split the data into 2 groups. Test to see if the average median household income in the youngest MSAs is lower than the median of median household incomes.

- What is the mean of median HH incomes for the older half of cities? Come up with a confidence interval estimate of the population mean.

- Do America's "young" cities have more rental properties than the average US city? Find the average number of rental units for all US MSAs and then test to see if the average number of rental units of the "young" cities is lower than the overall US city average.

- Will our tenants need dedicated high-speed internet access? How many people work from home in "young" cities? Find an interval estimate for the average number of people who work from home.

- Does the number of people who work from home in "young" cities differ from the national average? Test to find out.

- Draw any conclusions and make any recommendations you'd like to offer your client.

2. Retirement

- Does income vary with retirement? For this, define "retirees" as "people age 65 and up", and find the proportion of each city that is retirees. Then split the data into 2 groups. Test to see if there's a difference in household income between the MSAs with the most and least retirees per capita.

- Is a "work-from-home" attitude less common in retirement age cities? Do the proportions of citizens that work from home vary between "old cities" and other cities. Use the two groups to test if this proportion is lower in "old" cities than in other cities.

- Is there more of a vested interest at stake in the real estate market in "old" cities? Is the proportion of properties that are owner-occupied greater in "old" cities? Use these two groups to test for a difference between these proportions.

- Do retirees prefer smaller cities? Test to see if the total population is lower in "old" cities than in other cities.

- Is household size different in "old" cities? For each city, find the average household size and then test to see if this differs across the two groups.

- Draw any conclusions and make any recommendations you'd like to offer your client.

3. Education

- Are cities with college and graduate school enrollment more income-diverse or less so? For this, find the proportion of each city that is currently enrolled in college, graduate, or professional school. Then split the data into 2 groups

a. Test to see if there's a difference in average household incomes between the MSAs with the most and least proportions of the population in postsecondary education.

b. Test to see if there's a difference in the variance in household incomes between the MSAs with the most and least proportions of the population in postsecondary education.

- Do cities with college and graduate school enrollment have more rental property?

a. Test to see if there's a difference in the number of renter-occupied housing units between the MSAs with the most and least proportions of the population inpostsecondary education.

b. Test to see if there's a difference in the variance in the number of renter-occupied housing units across cities between the two groups.

- Do cities with college and graduate school enrollment have more high-earning households? For this, find the proportion of households in each city that earns $100,000 or more.

a. Test to see if there is a difference in the average representation of high-earning households between the MSAs with the most and least proportions of the population in postsecondary education.

b. Test to see if there's a difference in the variance in high-earning households between the two groups of cities.

- Draw any conclusions and make any recommendations you'd like to offer your client.

4. Rental Property

- For this, find the proportion of each city's housing stock that is currently renter-occupied.

- Use regression methods to address each of the following models:

- The percentage of the housing stock that is renter-occupied depends on

a. the percentage of the population that is young.

b. the percentage of the population that is retirement age.

c. the percentage of the population that is enrolled in college or graduate school.

- Which model is better between a, b, and c? What variables are significant? What is your interpretation?

- Household income depends on

a. the percentage of the housing stock that is renter occupied.

b. the percentage of the housing stock that is renter occupied and the housing stock per capita.

c. the percentage of the housing stock that is renter occupied, the percentage of the population that is retirement age, and the housing stock per capita.

- Draw any conclusions and make any recommendations you'd like to offer your client.

5. Some full models

- What are the determinants of household income? To examine this we will need to create some new variables:

Youth: define "young" as "between 14 and 24", and find the proportion of each MSA that is young Retirement: For this, define "retirees" as "people age 65 and up", and find the proportion of each city that is retirees.

Education: find the proportion of each city that is currently enrolled in college, graduate, or professional school

Rental intensity: find the proportion of each city's housing stock that is currently renter-occupied Housing stock per capita: the number of housing units per person in a city.

- Consider two models:

a. a model that posits that income depends on the retirement age (the best predictor of the above variables)

b. a model that includes youth, retirement, education, rental intensity, and the per-capita housing stock as independent variables.

• Does model a or b perform better? Use an F test for joint significance of a subset of variables to see which model should be preferred.
• Consider the residual plots. Are there any problems with these regressions?
• What are your interpretations of the preferred model?

- What are the determinants of the available housing stock per capita? Consider two models:

a. A model that posits that housing stock per capita depends on retirement.

b. A model that includes youth, retirement, education, and rental intensity as independent variables.

Does model a or b perform better? Use an F test for joint significance of a subset of variables to see which model should be preferred.

Consider the residual plots. Are there any problems with these regressions?

What are your interpretations of the preferred model?

Draw any conclusions and make any recommendations you'd like to offer your client.

Attachment:- Assignment File.rar