An Overview of the Methodology and Model Used to Estimate Long-Term Housing Surpluses and Shortages in Florida Using ACS and U.S. HUD Data

By Keith Ihlanfeldt, Ph.D. and Daniel Sierra

Prepared for the Housing Affordability Initiative
DeVoe L. Moore Center
Florida State University

March 28, 2025

Introduction

Florida is facing an unprecedented housing affordability crisis. Incomes simply have not kept pace with rising home prices. According to the U.S. Federal Housing Finance Agency, Florida’s home prices are seven times higher than in 1980. These home prices have outpaced the rise in national home prices steadily since 2016.

While the causes of rising home prices are complex, economists generally agree that one significant cause is the inability of supply to keep pace with demand. This is a particular problem in states like Florida, which have experienced rapid population growth due to migration from other states. While this internal migration is a source of pride for the state, the consequence is a need to add significantly more housing to accommodate these new residents and households.

Moreover, the severity of the housing shortage varies significantly among the state’s regions, metropolitan areas, and counties. This report represents the first systematic attempt to estimate housing shortages at the state and local levels. Overall, Florida faces a shortage of single-family homes and rental units of more than 120,000 units. This is a dramatic reversal from before 2010, when the state experienced an estimated excess housing supply of about 126,000 units (see figure below). Indeed, every county in Florida now experiences a housing shortage.

Over the next several months, the DeVoe L. Moore Center at Florida State University will publish profiles of estimated housing supplies for each Florida county. These profiles will also include recommendations for increasing the supply of housing so that counties can more effectively fit the rising demands of their local markets. The remainder of this section explains the methodology used to arrive at these estimates.

Methodological Overview

The two most widely used models for estimating housing shortages and surpluses are by Freddie Mac and Rosen Consulting Group. Freddie Mac is the Federal Home Loan Mortgage Corporation, a government-sponsored enterprise that plays a leading role in securing home mortgages and assisting in home finance. The Rosen Consulting Group (RCG) is a leading private sector real estate consultant.

The two models differ significantly in their structure, priorities, and assumptions they use for estimating housing needs. The Freddie Mac Model focuses on national housing trends, especially the downturn in entry-level single-family home construction. It uses historical production patterns and affordability issues faced by first-time buyers to estimate the amount of new housing that would be required to address an estimated shortfall. In contrast, the RCG model adopts a wider perspective on the supply-demand gap by comparing household formation rates and the implicit demand for housing, and housing stock declines against new construction. The RCG model frequently appears in policy debates regarding long-term national housing shortages and necessary investments.

Unfortunately, neither of these models was suitable for the current analysis because they cannot provide localized insights. The model developed by Florida State University economists Keith Ihlanfeldt and Daniel Sierra (K&S) and used by the DeVoe L. Moore Center improves on these models by explicitly incorporating vacancy rates.

Using five-year interval data from the American Community Survey (ACS), Ihlanfeldt and Sierra calculate the natural vacancy rates for rental and owner-occupied markets, offering deeper insights into localized housing dynamics. Unlike the Rosen and Freddie Mac methodologies, which primarily highlight national trends, the K&S model allows for the examination of unique regional and metropolitan variations, including those influenced by seasonal demand in vacation areas or by market pressures in urban centers. This level of detail is essential for developing targeted policy interventions to address specific housing imbalances rather than imposing generalized solutions. Additionally, calculating natural vacancy rates offers a dynamic measure of equilibrium against which changes in market conditions can be assessed.

The key differences between these methodologies lie in their assumptions and formulas for estimating housing shortages. The model established by RSG employs a formula that analyzes the demand-supply gap by comparing household formation and losses in the housing stock with new housing completions. While this method effectively describes broad national trends over the long term, it assumes a continuity of demand growth and risks overlooking localized imbalances in the housing market. Conversely, the Freddie Mac model, which primarily emphasizes the shortfall in constructing entry-level single-family homes, provides valuable insights into affordability challenges for first-time buyers; however, the market for new constructions is a relatively small sector. K&S utilizes variations in vacancy rates as an indicator of how tight the market is for housing, with “natural” vacancy rates serving as localized benchmarks for equilibrium. Higher vacancy rates would imply slack demand, while vacancy rates below the natural rates would imply more competitive housing markets. This approach is more adept at capturing real-time instances of housing imbalance and attending to the nuances of local markets.

The K&S methodology is an adaptation of the National Homebuilders Association’s model (NAHB, 2022) to estimate housing shortages at the national level. The methodology involves comparing observed vacancy rates to an estimated natural vacancy rate and using the resulting differences to estimate housing shortages in numerical terms. The “long run” average vacancy rate serves as a proxy for normal, or natural, vacancy rates.

Estimating Housing Shortages

Calculations and Definitions

K&S’s methodology relies on several key variables for each county and five-year period (i):

ForRent_i: The number of units available for rent during period i.
RenterOcc_i: The number of rented and occupied units during period i.
RentedNotOcc_i: The number of units rented but not currently occupied during period i.
ForSale_i: The number of units available for sale during period i.
OwnerOcc_i: The number of owner-occupied units during period i.
SoldNotOcc_i: The number of units sold but not currently occupied during period i.

Renter and Owner Vacancy Rates

The renter vacancy rate is calculated as the proportion of rental units available for rent to the total number of rental units in the market. Similarly, the owner vacancy rate represents the proportion of homes available for sale to the total number of owner-occupied housing units.

RenterVacRate_i = ForRent_i / (ForRent_i + RenterOcc_i + RentedNotOcc_i)
OwnerVacRate_i  = ForSale_i / (ForSale_i + OwnerOcc_i + SoldNotOcc_i)

Natural Vacancy Rates

The natural vacancy rate is calculated as the long-term average vacancy rate for each county, representing a balanced market where supply meets demand. Here, n is the total number of five-year periods (2006–2010 through 2018–2022), equal to 13.

NaturalRenterVacRate = (Σ_i=1..n RenterVacRate_i) / n
NaturalOwnerVacRate  = (Σ_i=1..n OwnerVacRate_i) / n

Vacancy Rate Difference

The vacancy rate difference quantifies the deviation of the observed vacancy rate from the natural vacancy rate. A positive difference indicates that the observed vacancy rate is lower than the natural rate (shortage). A negative difference implies a surplus.

RenterVacDiff_i = NaturalRenterVacRate − RenterVacRate_i
OwnerVacDiff_i  = NaturalOwnerVacRate  − OwnerVacRate_i

Housing Shortage (Units)

The numerical housing shortage is computed by multiplying the vacancy rate difference by the total market size for each tenure.

RenterShortage_i = RenterVacDiff_i × (RenterOcc_i + RentedNotOcc_i + ForRent_i)
OwnerShortage_i  = OwnerVacDiff_i  × (OwnerOcc_i  + SoldNotOcc_i  + ForSale_i)

Cumulative Housing Shortage

The cumulative housing shortage for each period is the sum of renter and owner shortages:

CumulativeShortage_i = RenterShortage_i + OwnerShortage_i

Case Application: Alachua County (2018–2022)

In Alachua County during the 2018–2022 period, the renter vacancy rate was calculated as the ratio of units available for rent to the total rental market size, including rented and occupied units, rented but not occupied units, and units available for rent. The owner vacancy rate was similarly calculated as the ratio of units available for sale to the total owner housing market size, including owner-occupied units, sold but not occupied units, and units available for sale.

The natural vacancy rates for renters and owners were determined as the average of observed vacancy rates across all 13 periods. These natural rates were then used to calculate the vacancy rate differences for 2018–2022. The renter vacancy difference was the natural renter vacancy rate minus the observed renter vacancy rate for 2018–2022, while the owner vacancy difference was the natural owner vacancy rate minus the observed owner vacancy rate for the same period.

Using these differences, the renter housing shortage was computed by multiplying the renter vacancy difference by the total rental market size, and the owner housing shortage was computed by multiplying the owner vacancy difference by the total owner market size. The cumulative shortage was then obtained by summing the renter and owner shortages. For example, if the renter vacancy difference was positive, this indicated a shortage requiring additional units to restore balance. Conversely, a negative difference would indicate a surplus.

For Alachua County, the estimates show for the 2018–2022 period a shortage of owner-occupied housing units equal to 581, and a surplus of rental units equal to 989. With few exceptions, counties throughout the state are experiencing shortages in both types of housing that have worsened over the past decade. As expected, these shortages are greater in counties with larger housing stocks.

Recommended citation

Keith Ihlanfeldt and Danny Sierra, “An Overview of the Methodology and Model Used to Estimate Long-Term Housing Surpluses and Shortages In Florida Using ACS and U.S. HUD Data,” Prepared for the DeVoe L. Moore Center Housing Affordability Initiative, College of Social Sciences and Public Policy, Florida State University, March 28, 2025.