Mapping America's Rent Crisis

This is my final project for PP434: Automated Data Visualization for Public Policy. My assingment is to make 5-8 data visualizations on a policy-related topic of my choice and discuss my aims, data, analytical challenges, and conclusions.

Aims

This project aims to (i) visualize the scale of the American housing affordability crisis and (ii) identify regional trends in both housing supply and housing demand.

Rent is Expensive

The United States is facing a once-in-a-generation housing affordability crisis. In mid 2022, a typical American renter devoted a greater share of her income to rent than at any point this century.

The map below shows median gross rent as a share of household income in every US county. Households that pay more that 30% of their income on rent are officially considered "rent burdened". A typical renter is near or above this threshold in all of America's population centers.

I accessed the data above through the US Census API. I wrote a python function to form a time series rather than a single year's data. A secondary goal for this project was to reflect the true population density of the US while showing regional trends. US population density doesn't lend itself well to straightforward choropleths, so I find cartograms to be more informative. 1

Rents surged in the wake of the COVID-19 pandemic as surging demand met chronically low supply. The crisis peaked in mid-2022, with rents growing nearly twice as fast as overall prices.

I decided to use Zillow's Observed Rent Index (ZORI) for this project. Zillow has monthly data stretching back nearly ten years and they calculate the figure such that it doesn't over-index on new housing. Zillow doesn't have a free API, so I had to harvest stable links from Zillow's site. I accessed inflation data using FRED's API.

To understand the American housing market, it helps to understand where people are moving. New York, Illinois, and California – home to the three largest cities in the US – are hemorrhaging population, while the south and mountain west have seen large gains. Housing costs are a major driver away from major coastal cities and toward the "sun belt." Population loss is a fate all areas of the country should seek to avoid. It means fewer jobs, less tax revenue, and less political power. 2

Mapping Rent Inflation

The first map below shows the change in ZORI over the past five years by county. Few counties have been immune from spiking rents. The southeast and southwest saw the greatest increases in rent and their populations surged. Florida seems to be the most impacted state.

The one-year view tells a different story. Over the past year, rent inflation is felt most acutely in the midwest and northeast. In the sun belt, rent inflation has leveled off. A small handful of counties, mostly in Texas, even saw their rent index decrease over the past year.

One limitation of the Zillow data: ZORI is fairly robust in its coverage, but several smaller counties are missing from these cartograms. To appear on the map, a county must have data for the most recent period and one or five years prior.

Supply and Demand

Zillow also has a variable called the Zillow Observed Rent Demand Index (ZORDI). It measures the traffic on a typical listing in an area. Think of it as a barometer of competition for available units. The visualization below shows ZORDI and the year-on year change in ZORI in America’s 50 largest metro areas.

The plot sorts itself into regions remarkably well. In the southeast and southwest, prices are levelling off as renters face less competition. In the northeast and midwest, competitive markets are seeing the steepest rent increases.

This is a supply-driven market dynamic. The cities and regions with the highest building rates have seen the slowest rent growth. Using data from the US Census Bureau's Building Permits Survey, I calculated the number of new multi-unit housing structures approved per capita in the past year. I chose multi-unit structures because they are usually built to rent.

Keep in mind that this is new housing approvals, not completions. Just the knowledge that new housing is on the way can be a signal to landlords that they must price competitively. The city of Austin, Texas breaks this chart. No city has built more housing and no city is seeing rents decrease faster. This is remarkable considering Austin is one of the fastest-growing cities in the country.

How's Florida Doing?

The maps above show Florida is the most rent-burdened state in the country and the state with the highest rent inflation since the pandemic. The sunshine state has also gained the most population in relative terms. So it's worth taking a closer look at the state. The charts below show homebuilding and rent growth in Florida's four major metro areas, which are home to more than half of its population.

Orlando and Jacksonville have built more homes than the rest of the state and have seen their rent growth slow as a result. Rents are still extremely high, to be sure, but they have leveled off. If these cities's building rates keep pace, incomes should eventually catch up to rents and the cities will become more affordable.

Conclusion

The above visualizations demonstrate the need for abundant housing in America. The cities and regions that have built the most housing have seen the slowest rent growth. The cities that have built the least have seen the most. After an instense spike, there is reason for optimism, but only in some areas.

In southeast and southwest, building lots of new housing has paid off and rent growth has stabilized. If the midwest and northeast do not build more housing, they will continue to become more expensive and lose more population. This is a fate all areas of the country should seek to avoid.

Data Visualization Portfolio
Diverging Life Expectancy