Assessing the Affordable Housing Gap

Housing systems are complex, with several factors affecting why people live in their current housing. This housing data story explores the affordable housing gap through the lens of supply, demand and the dynamics within the housing system to paint a more complete picture of the region’s gap in affordable housing. 

At a basic level, the total number of housing units in the region is greater than the number of households, given that some units are vacant  ¹  . But as we know, not all units are priced the same and not all households have the same income to spend on housing.

There are many reasons householders live in the housing units they occupy. Affordability is a major factor. The U.S. Department of Housing and Urban Development (HUD) uses 30% of gross income to  define  an affordable housing cost. 

Using this definition, we can match households by income to an affordable housing unit by price point, measured in terms of  monthly housing costs . This begins to paint the picture of the housing gap we have in the Kansas City region  ²  . When we see fewer housing units at or below a price point than the number of households that need housing at that price point to stay within an affordable housing range, we call this an affordable housing gap. 

The gap primarily impacts renters and does not appear to be as significant for those who currently own their home. This does not take into account renters who would prefer to become homeowners and the barriers they face, which may include affordability. 

The largest affordable housing gap is for households looking to rent units for less than $650 per month. The chart below shows a gap of 45,449 units at the regional level when looking at housing units by price point compared to households by income. The gap shown at the high end is not an affordability gap, as these households can afford any of the housing in the region. 

The magnitude of the renter affordable housing gap varies throughout the region and is clearly shown when broken down by county. While Jackson County faces the largest gap by number of units (23,146), Johnson County faces the largest gap by proportion of households. Johnson County’s gap is more than three times the level of its existing rental housing stock at this price point (2,968 units). Meanwhile, Ray County appears to have no shortage of affordable housing.

This basic gap analysis provides quick insight into the mismatch of the number of households in certain income ranges compared to the number of housing options available to them. 

However, this calculation doesn’t consider the dynamics within the housing system of lower-income households affected not only by the total unit shortage, but also the reality that some higher-income households occupy units affordable to low-income households.  

Exploring this interconnected dynamic shows a likely wider affordable housing gap but requires a different data set to understand. Comprehensive Housing Affordability Strategy (CHAS) data groups households by their income relative to the metropolitan area Median Family Income ( MFI ) and groups housing units by which household income group they are affordable to. As a result, you will see price points represented using these relative income groups.  For ease of understanding, we also show the approximate income levels and affordable rents consistent with these groups in the table below.

Using this data, we can see the rental housing units that are affordable to each income group combined with which income group the units are actually occupied by. We continue to focus on renters because that is where the affordability gap is most severe  ³  . The table below represents the number in each category within the nine-county Kansas City region.  

A Sankey diagram best depicts the dynamic, showing households by income group on the left compared to units by the income group they are affordable to on the right. From the diagram, it is easy to see that simply because housing units are affordable to households with certain income levels doesn’t mean those households are the only, or even the primary income group living there.  

Breaking the gap down by county like we did when matching household income to housing unit price point, it is evident that the severity of the renter affordable housing gap still varies, although less dramatically, throughout the region. When accounting for crowding, we can see that Ray County does, in fact, have unmet housing needs for ELI renters (a group that is roughly comparable to renters needing units for less than $650 in the previous section). 

The data highlighted in this data story demonstrates there is no single or simple solution to the affordable housing challenge. As the region continues to build strategies to meet its current and future residents’ needs, this assessment provides an additional lens for context and informed conversation. 

The assessment helps inform solutions in two ways. First, it establishes that the region has a gap of about 64,000 affordable housing units. Second, it identifies the market segment in which this deficit is most acutely felt — renters with extremely low incomes who can afford rents below $650 per month. The market alone cannot supply traditional new units at this price point. As a result, expanding supply will require substantial public subsidy, new housing types or a combination of the two.    

This data story also illuminates a complication of increasing the supply of affordable housing for those with the lowest incomes. People with higher incomes may choose to consume this supply if given the opportunity. Most households spend substantially less than 30% of their incomes on housing when given the choice. What is technically affordable to VLI or ELI renters may not be available to and attainable by them.  

Existing homeowners, on the other hand, have a sufficient stock of housing that is affordable to, available to, and attainable by them based on this analysis. Other criteria, like housing quality or desirability of the location, may be more significant contributing factors to owner dynamics than supply and demand. The next data story will start to explore the roles and realities of housing stock location in the Kansas City region.  

• Data Limitations: ACS data used in this data story is updated through 2021. CHAS data is only available through 2019 at the time of this publication and is therefore notably lagged. Rents have increased much faster than incomes since 2019 and interest rates have doubled. As a result, the true affordable housing gap in 2023 is likely larger than the one calculated using 2019 data.  
• Forecasting: The data used in this analysis does not include forecasting. The dynamics demonstrated in this data story will certainly be impacted as the region looks to attract new businesses with a significant number of employees to the region. The region also  anticipates the loss of some traditionally affordable Low-Income Housing Tax Credit (LIHTC) program rental units  in the absence of efforts to preserve this affordable housing stock.  
• Housing quality: This analysis does not consider housing quality or location desirability, which may give more insight into understanding affordability issues for owners. A unit may be affordable to, available to, and attainable by a household looking to purchase, but still not meet some minimum criteria based on condition or location. 
• People experiencing homelessness: The census definition of households does not include people experiencing homelessness. Therefore, estimates using households as a proxy for demand are an underestimate of the actual need.  
• Short term rentals: Short term rentals are primarily counted as  vacant  by the Census Bureau, but these units are not actually available to households looking for a unit to rent long term. Therefore, the stock of available units used for this analysis is likely an overestimate because some may be non-owner occupied STRs.
• Spatial mismatch: This analysis does not take into account the spatial mismatch between where people want to live and the stock of affordable housing. Spatial mismatch will be the topic of a future data story. 
• Specific recommendations: Minimal conclusions can be drawn about the composition of the housing stock that is needed outside of the price points and/or income groups provided in the analysis. 
• Transportation: Access to transportation is not considered in this analysis. Affordable transportation options can help minimize the financial strain caused by housings costs for some households. 

Note: A previous  data story  presented an estimate of “over 70,000 households with unmet affordable housing needs.” This number differs from those given in this data story because 1) it was based on an earlier version of ACS data and 2) it was calculated using all households making less than 20,000 per year compared to rental housing units costing $500 or less per month.  

American Community Survey (ACS) 5-Year Data:

Tables used for Housing Units:

• For Renters
  • Table B25063: GROSS RENT, American Community Survey 2017-2021 (occupied)
  • Table B25061: RENT ASKED, American Community Survey 2017-2021 (vacant)
• For Owners
  • Table B25104: MONTHLY HOUSING COSTS, American Community Survey 2017-2021 (occupied)
  • Table B25085: PRICE ASKED, American Community Survey 2017-2021 (vacant): Monthly costs based on assumed loan to value ratio of 70% and mortgage terms of 6.5% for 30 years. Utilities and other costs of ownership are not included in the approximate monthly housing cost, which is based on the value of the home.
• Unless otherwise noted, utilities are included in rent and housing costs.

Tables used for Households:

• Table B25118: TENURE BY HOUSEHOLD INCOME IN THE PAST 12 MONTHS (IN 2021 INFLATION-ADJUSTED DOLLARS), American Community Survey 2017-2021

Where income ranges and price points did not align, apportionment was used to allocate counts to the necessary ranges.

Comprehensive Housing Affordability Strategy (CHAS) Data

Table used for affordable to and occupied by:

• For Renters:
  • Table 15C: Substandard Housing by Renter Affordability by Household Income Bedrooms (explicitly includes utility costs)
• For Owners:
  • Table 15B: Substandard Housing by Renter Affordability by Household Income Bedrooms (does not explicitly include utility costs)

The CHAS data, provides data on households and housing units grouped by income categories as percent of the metropolitan area Median Family Income (MFI). Importantly, incomes are adjusted based on family-size in a manner similar to the 2021 income limits shown here. Larger households have greater non-housing expenses, and so need lower rents for a unit to be affordable. 

Comparing CHAS and ACS Data: Income and rent dollar amount vary by family size using this CHAS data, which shows housing affordability and income group as a percent of the dynamic MFI based on family size (see table above). 

ACS data is based on a dollar amount for income and a dollar amount for housing costs (using monthly ownership costs, housing price asked, gross rent, and rent asked). The affordability amount is an approximation for this dataset.  

We use both data sources in this story for multiple reasons. First, ACS data is updated more quickly than CHAS, and more people are familiar with ACS data. We use ACS data for a lot of the dashboards available on the Housing Data Hub. However, ACS data does not have the dynamic affordable to and occupied by data that CHAS offers, which is crucial to understand crowding and the true size of the affordable housing gap. 

Click and hover to interact with the Sankey diagrams below.

1. By definition, the number of households equals the number of occupied housing units. A housing unit occupied at the time of interview entirely by people who will be there for two months or less is classified as "Vacant-Current Residence Elsewhere". Such units are included in the estimated number of  vacant units .
2. Kansas City region refers to the following nine counties: Cass County, MO; Clay County, MO; Jackson County, MO; Johnson County, KS; Leavenworth County, KS; Miami County, KS; Platte County, MO; Ray County, MO; Wyandotte County, KS.  

Please use the form below to submit any feedback for this story!