Housing Landscape

This interactive web application combines maps with narrative text, images, and multimedia content. Given its design, you need only scroll down to explore this report.

Each map is interactive, allowing you to explore more than the default visible features. To interact with them, follow these instructions:

Legend button: Clicking the icon on the lower left corner will show the legend of the currently visible map layer. This will show all the icon and color descriptions needed to understand the map. To exit the legend, click the close button in the upper right corner.

Expand button: This icon, in the upper right corner, will display the map in full-screen. Particularly useful for viewing the details of smaller maps. To exit full-screen mode, go to the minimize button, located in the same place.

Zoom in/out and Home buttons: Located at the lower right corner, these icons are particularly useful for maps where you want to see small details of particular areas or when the map's elements overlap each other. For Swipes (maps put next to each other to be compared using a swipe view), zoom-ins and zoom-outs on one map will affect the other, and the lower left corner Legend button opens the legend of both maps. Click on Home to go back to the default view. These buttons can be used after Expand to increase the detail in the view you want.

Pop-ups: Most polygons (colored shapes representing census tracts) on the maps include extra information that can be accessed by clicking on them. When clicking small overlapped polygons, a page selector appears in the pop-up's lower right corner, allowing you to see the details about all the elements selected.

Housing constitutes an essential component of citizens' well-being and quality of life. Knowing and mapping the state of the housing market is crucial to understanding current conditions and formulating effective public policies. In mapping such broad geographic regions, the availability of granular data becomes essential, as overall averages can hide critical details only revealed through disaggregated, geographically specific analysis. This report uses a tract-based approach that allows the identification of local disparities and needs, providing public policymakers with insights to plan equitably and efficiently. In this context, the present study on housing trends in Alabama, Louisiana, North Carolina, and Tennessee becomes a helpful tool for better understanding the challenges and opportunities faced by the different regions within each state, promoting balanced and sustainable development.

The study addresses four main questions:

1. What is the current state of housing in Alabama, Louisiana, North Carolina, and Tennessee?
2. How has economic and population growth impacted the housing market?
3. How has urban growth affected housing demand for and affordability of rural housing?
4. Who are the residents most affected by recent trends?

Analyzing the housing market can become a tremendously complex task given the multiple variables, indicators, metrics, and sources of information that can be explored. For this study, the American Community Survey (ACS) 5-Year Estimates for 2012, 2017, and 2022 were used, particularly table DP04, which focuses on housing. Since these estimates add data from the year in question plus the previous four years, using these three points in time makes it possible to see a progression over 15 years (between 2008 and 2022). Another advantage of using this data is that it allows for the desired geographical disaggregation by offering data at the census tract level for the periods in question.

For this first question, this study synthesizes some key market variables, namely, total units, separated between vacant and occupied. The occupied ones are divided between owner-occupied and renter-occupied. From this last classification, the owner-occupied were split between those with and without a mortgage. Of the renter-occupied, the % of units based on the proportion that rent represents of their household income is indicated. Additionally, for the owner-occupied, the tract average of the median value of the houses at the tract level is shown. The median gross rent payment, as tract average, is shown at the tract level for the renter-occupied.

While Table 1 summarizes the situation at the state level, showing that Alabama (closely followed by Louisiana) has the highest vacancy rate and homeownership rate and that Tennessee has the lowest vacancy rate, the map reveals the complexity of the data.

Considering that the definition of tracts addresses a certain number of statistically relevant populations, it is clear that the size of the tracts will have a tremendously high variation. The map below, on the left, shows the distribution of total units and, on the right, the unit density (units per square mile). On both maps, the areas that, according to the 2020 decennial census, are considered urban can be seen with black borders.

If you click on any tract (on this and any following map), the pop-up table will summarize all the data in Table 1.

This contrast shows that, when counting total units, there are tracts with a large size that, far from urban areas, have a high value. Still, the value is reduced when standardizing it to the density of units, leaving these densely populated tracts restricted to urban areas and their immediate surroundings.

While Table 1 shows that occupancy is between 80 and 90%, the following map shows that, as expected, tracts with large amounts of vacant units are geographically concentrated outside urban areas.

Another ACS variable provides relevant insights: the age of the housing stock. The survey counts units according to the period in which they were built. Using 5-year estimates from the 2022 ACS, the following table summarizes the situation at the state level:

The data granularity also enabled the creation of the following map showing the geographic distribution of the housing stock by age. The map shows one point per 500 units and follows the color code described in the legend. A typical pattern in the age distribution of housing units, particularly in large urban areas, is a concentration of older units at the center and newer developments growing around them.

Limiting the analysis only to occupied units, these are classified between owner-occupied and renter-occupied. Table 1 shows that the owner-occupied ratio is above 65% at the state level for all four states, with Alabama being the largest (69.7%). However, this percentage is not evenly distributed geographically, and, as the following maps show, the ratio of renter-occupied units is exceptionally high within and around urban areas.

The ACS divides owner-occupied units into those with and without a mortgage. Table 1 shows that having a mortgage more than triples the selected monthly owner costs, which are exceptionally high in North Carolina.

Another relevant variable when analyzing owner-occupied units is their value. According to Table 1, the median home values ​​differ among the states, particularly between Alabama (the lowest median value) and North Carolina (the highest), with more than $75,000 difference in the 2018-2022 averages.

The density graph on the right, created with tracts as the unit of analysis, shows that this difference goes beyond the average and is not caused by a few exceptionally high outliers. On the opposite, the whole distribution of North Carolina is right-shifted compared to Alabama, meaning the Alabama tracts have consistently lower home values ​​than the North Carolina tracts.

The following map shows the spatial situation of the previously discussed variables: owner-occupied units with mortgages and the average value of said units. Although it seems urban areas correlate with having a mortgage and higher median value of houses, there is more dispersion than other variables reviewed previously (especially regarding having a mortgage, which is pretty common in many areas outside the urban).

When analyzing renter-occupied units, an important concept is cost burden. Generally, households spending more than 30% of their income on housing costs are considered cost-burdened. According to the breakdown in Table 1, the proportion of households that rent alone represents more than 30% is around 50% in all states, and it’s exceptionally high in Louisiana (54%). Note that this metric underestimates the cost burden, as it omits other expenses related to housing and focuses solely on rent payments.

The other relevant variable to study in this section is gross rent. According to Table 1, the state average of the tract median gross rent is higher in North Carolina. As with the median home value, this result is not due to outliers but to a distribution of median income at the tract level consistently higher than other states.

The maps below show both variables. While higher income levels are predictably correlated to urban areas and their contours, the story with the cost burden proxy is different. Although rent is higher in urban areas, so are incomes; the cost-burdened category can apply to many other regions outside the urban areas that, although with possibly lower rent levels, face much lower incomes.

The first step is to measure economic and population growth. Table 3 shows the population growth of the four states, with North Carolina and Tennessee as exceptional cases. Between the ACS for 2008-2012 and 2018-2022, there was an increase of 10%.

In economic terms, the average median income is presented at the tract level for families and households (housing units that may be made up of unrelated people), expressed in 2022 dollars after being corrected for inflation. For all states, household income is higher than family income, and it also shows a slightly higher percentage increase, always with North Carolina in the lead.

It is important to understand how the housing market evolved during these same periods. Table 4 presents the same variables as Table 1, summarizing essential metrics for the housing market, this time for the three periods under study.

Analysis of the data indicates:

• While the total number of housing units, both owner-occupied and renter-occupied, increased during this period, the proportion seems stable (~70% vs 30%) over time.
• The percentage of owner-occupied units with mortgages decreased in the four states.
• The percentage of renter-occupied units with a rent representing more than 35% of income decreased generally, except in Louisiana, where this number was maintained.

Given that the  Housing Affordability Index  calculated by the National Realtor Association (NAR) was not available at the tract level, nor were the years under consideration in the report, following their approach, the formula on the right compares median home value to median family income to calculate a price-to-income ratio. The higher the ratio, the lower affordability is, meaning a median family is less likely to afford a median-valued home.

Table 5 shows the state-level value of this ratio for the four states and the three temporal points. It shows that Louisiana's ratio has consistently increased, meaning its affordability has consistently decreased. On the other hand, while Alabama has the lowest ratio in general, it faced the same pattern as North Carolina and Tennessee: an increase between 2012 and 2017 but a decrease between 2017 and 2022.

At the tract level, these state aggregate values ​​have much more dispersion. As the map below shows, affordability is negatively related to urban areas, consistent with the high home values ​​in those regions. Another conclusion drawn from the map is that tracts with high affordability (i.e., low price-to-income ratio) are more common in Alabama and Louisiana, consistent with the state averages shown in Table 5.

Homeownership is another relevant metric worth analyzing urban areas against, which can be estimated with the owner-occupied unit’s rate. A more granular approach to evaluating this relationship can be achieved by calculating the percentage of the population that, according to the Decennial Census 2020, is defined as urban. This allows the "level of urbanity" to be a continuous value between 0 and 1 for each tract and not a dichotomous variable based on geographies of urban areas that do not coincide with the tracts.

The previous scatterplot matrix shows the relationship between this urban population rate (UPR) and the owner-occupied unit rate. As can be seen, higher percentages of the population considered urban are related to lower levels of homeownership, confirming that, at the tract level, urban contexts are negatively correlated with affordability and homeownership.

Finally, Table 6 shows how these relationships with homeownership and affordability are maintained when the analysis is restricted to tracts defined as fully rural or fully urban (based on the ACS definition of urban population).

With levels of affordability and homeownership rising in some states and periods and falling in others, and with much more variation at the tract level depending on their proximity to urban centers due to heterogeneous increases in both income and house values, it is not easy to isolate all these effects to define which populations should be focused on to answer this question.

For this section, the 10% least affordable tracts were selected, and the demographic data of the population living in them was analyzed.

Considering only the tracts for which the two components of the affordability index (median home values ​​and median family income) were available, the 10% with the highest price-to-income ratio are 684 tracts with a ratio above 3.89, as shown in the following map. The map also displays the areas defined as urban by the 2020 Decennial Census.

More than 70% of the selected tracts are in North Carolina and Tennessee (281 in North Carolina, 203 in Tennessee, 148 in Louisiana, and 52 in Alabama). Below are the demographic and economic data of the populations in these tracts:

Analysis of the data indicates that the people living in these tracts are:

• While for Alabama, North Carolina, and Tennessee, the majority of the population is white; for Louisiana, the majority is Black, closely followed by the white population.
• Between one-fifth and one-fourth of the population is under 18 years old, except for North Carolina, the only state with a median age above 40.
• The population has an unemployment rate of around 5%, and around 20% of people are in poverty, except for Louisiana, where unemployment is more than 50% higher than the rest, and more than 1 in 4 people is in poverty.

This study on the housing market in Alabama, Louisiana, North Carolina, and Tennessee provides relevant insights into these regions' unique challenges and dynamics. Detailed analysis of disaggregated data at the census tract level reveals patterns and trends not visible in state-wide averages, highlighting the importance of granular data for informed policymaking.

Key findings from this review include the strong correlation between the density of housing units and occupancy rates within urban regions. North Carolina has the youngest housing stock, while Louisiana has the oldest. Similarly, other variables are significantly differentiated by state; for instance, the median home value and the median gross rent are highest in North Carolina and lowest in Alabama. Conversely, some variables remain relatively stable across states but show significant differences between urban and rural contexts, such as homeownership rates compared to renter ratios and levels of affordability.

Additionally, the study measures economic and population growth, which, although varying in magnitude, was experienced by all states. This growth led to a decrease in the percentage of owner-occupied units with mortgages and renter-occupied units spending more than 35% of their income on rent. However, it also resulted in reduced affordability levels (meaning that the median home value increase outpaced the median family income); meanwhile, the ratio between owners and renters remained relatively stable.

Finally, after calculating a price-to-income ratio, the study identified regions where home purchasing is less likely. These regions were then analyzed demographically, revealing that they are primarily located in North Carolina and Tennessee, states that feature tracts with low affordability that have higher income levels and lower poverty rates.

In terms of potential use beyond these findings, the extensive data available in the interactive maps serve as a powerful tool for understanding and contextualizing the housing conditions across numerous regions. These maps enable readers to see beyond averages and tailor strategies to address specific local needs, promoting equitable and sustainable development, and demonstrating the value of detailed, geographically specific data in effectively addressing housing market challenges.

Below is a link to an interactive mapping portal where readers can create their own maps based on the data found throughout this StoryMap. We encourage users to explore the data for themselves using the following button.