Forecasting the Demand for Housing in the United States

Forecasting the Demand for Housing in the United States

G. Stacy SirmansFlorida State University

David A. MacphersonTrinity University

Forecasting Housing Demand

A Research Project Sponsored by the

REALTOR University Research Center

National Association of Realtors

March 2012

Introduction

Interesting time for housing demand in

U.S.

2000s decade: Real estate boom; now

working through period of correction

Simultaneously, baby boomer generation is

entering retirement

Should reverberate through all sectors of

economy including the housing market

Introduction

BIG Question:

What is the future of U.S. housing

demand given these factors?

Smaller Questions:

What will be baby boomer living

arrangements?

How will younger generation absorb

housing inventory?

Introduction

This study uses a multinomial logit

model to forecast the demand for

housing in the U.S.

Part One estimates U.S. housing

demand

Part Two estimates demand for the ten

largest states

Population

The level and makeup of population set

the base for housing demand

All age categories of adult population

increase in absolute numbers, but all

decrease as a percentage of total except

the age group 65+

65+ age group increases 78% from 17%

to 26% of adult population

Population

Population

Modeling U.S. Housing Demand

Model segments housing into five

categories:

Owned Detached Housing

Rented Detached Housing

Owned Attached Housing

Rented Attached Housing

Owned/Rented Mobile Home


Multinomial logit model is used to identify the

determinants of housing demand

Dependent variable relates to these housing

choices:







The explanatory variables include

gender, age, and race

The estimated coefficients are converted

into partial derivatives, measuring the

impact of a one-unit change in the

explanatory variable on the probability

of being in a given housing category


The partial derivatives are combined with the

Census population projections by gender,

race, and 5-year age group to generate a

forecast of the demand for the five housing

categories

Housing forecasts over 2010-2030 are

provided for the U.S. overall and for all 50

states, with discussion of the 10 largest states


The primary data source is the 1%

2007-2009 American Community

Survey Public Use Microdata Sample

The sample is restricted to U.S.

residents age 18 and over

The sample size is 6,693,821


Housing stock depreciation rates come

from the 1999-2009 HUD Components

on Inventory Change Reports

State-specific vacancy rates are from the

U.S. Census Bureau Housing Vacancy

Survey Reports

Results for U.S. Overall



Largest category (54%)

Number of units demanded expected

to grow about 20%

Proportion of total housing

demanded remains about the same



Forecast to decrease slightly as

proportion of total housing

demanded

Number of units demanded expected

to grow about 17%



Forecast to experience the largest

increase (32%) in housing demanded


demanded expected to increase

slightly from about 7.62% to 8.35%



Second largest category of housing

demanded and is forecast to increase

21%


demanded expected to increase only

slightly to 23%



Smallest housing category demanded

and forecast to remain about 6% of

total

Total units demanded will increase

about 20%


A white male age 55-64 is the most likely

of all categories to demand Owned

Detached Housing

The likelihood of demanding Rented

Detached Housing decreases with age

for all categories


Non-white males and females in all age

groups are more likely to demand

Owned Attached Housing than either

white males or white females

Non-white females age 18-24 are the

most likely of any category or age group

to demand Rented Attached Housing


White females age 18-24 have the

highest probability of all categories and

age groups of demanding a mobile home

Non-white males and non-white females

of all ages have a lower likelihood of

demanding a mobile home than white

males age 18-24


Non-white males and females age 40-54

are less likely to demand Owned

Detached Housing than their white

counterparts

Non-white females age 40-54 have the

highest probability of demanding

Rented Detached Housing than all other

age groups or categories


White males age 40-54 are the least

likely to demand Rented Detached

Housing for all ages and categories

Starting at age 40, white females are

more likely than white males to demand

Owned Attached Housing and the

probability increases with age


Non-white females age 65+ are the most

likely to demand Owned Attached

Housing than all other categories and

age groups

The likelihood of a white male or white

female demanding Rented Attached

Housing decreases with age until age 65

Results by State - CaliforniaOwned Rented Owned RentedMobile

Detached Detached Attached AttachedHome

2008 Units 5,699,015 1,460,906 918,096 3,651,295451,799

2030 Units 7,146,934 1,754,602 1,212,908 4,468,672596,687

% Change 25% 20% 32%22% 32%

Proportions

2008 46.79% 11.99% 7.54% 29.98%3.71%

2030 47.16% 11.53% 7.99% 29.37%3.95%

Results by State - California

Results by State - Florida

Owned Rented OwnedRented Mobile Detached

Detached Attached Attached Home

2008 Units 3,500,478 602,115 875,3541,418,235 632,014

2030 Units 5,454,939 893,749 1,563,0082,148,330 1.038M

% Change 56% 48% 79%52% 64%

Proportions

2008 49.81% 8.57% 12.45%20.18% 8.99%

2030 49.15% 8.05% 14.08%19.36% 9.35%

Results by State - Florida

Results by State - Ohio

Owned Rented Owned RentedMobile Detached Detached

Attached Attached Home

2008 Units 2,788,250 415,261 200,070 949,816 165,557

2030 Units 2,822,172 409,292 234,634 966,860 165,253

% Change 1% -1% 17% 2%-0.2%

Proportions

2008 61.70% 9.19% 4.43% 21.02%3.66%

2030 61.38% 8.90% 5.10% 21.03%3.59%

Results by State - Ohio

Results by State

Although each state has its own growth

pattern, some common themes emerge:

The largest growth categories are Owned

Detached and Owned Attached Housing

The northern states (Illinois, Pennsylvania,

Ohio, Michigan) are predicted to have

relatively low increases in housing demand

Results by State

Although each state has its own growth

pattern, some common themes emerge:

The southern states (Texas, Florida,

Georgia, North Carolina) are predicted to

have much greater future housing demand

Although most states will experience an

overall increase in housing demand, the

proportionate demand for each category is

expected to remain relatively constant

THE END

Forecasting the Demand for Housing in the United States

Business