Urban Housing Markets
Prof. Clark
Week #12
Urban Housing Markets
Brief summary
Spatial patterns in the housing market
Recall insights of SUM regarding K-Land ratio
as you move away from place of employment
Look at some of the characteristics
associated with the housing good
Characteristics of the Housing
Good
Uniqueness
Features, locational fixity.
There are neighborhood effects
Positive and negative externalities
Long-lived
Stock of housing generates a flow of services.
Housing deteriorates over time without
maintenance.
Consumption and investment good
Has been seen as a hedge against inflation
• More later
Characteristics - continued
Housing is typically the largest component
of the household budgets
Requires financing.
There are tax advantages to owning housing.
Low income typically rent, whereas middle and
upper class typically own.
Market Structure
Monopoly power on supply side of
housing market?
Monopolistic Competition
Why?
Housing market is segmented.
What does this mean?
Housing Market Analysis
Use simple supply and demand analysis of
market.
The stock of housing generates a flow of
services to the consumer.
The stock of housing sells at the asset price,
V.
The flow of services sells at the flow price, R.
The two prices are related by the user cost of
housing, R=*V
User Cost of Capital
Defined
User cost represents costs associated
with the holding of housing capital.
Think of this as the costs to the owner of
housing.
Costs include the sum of mortgage
expenses, maintenance, and property taxes,
less any capital gains.
User Cost of Housing
Capital
Components of
nominal mortgage rate (i=r+ e) which is
deductible at the federal income tax rate (t),
property tax rate (p) which is also
deductible, maintenance expense rate (m)
which increases with age of property, capital
gains (g=gr+ e)
[(r+ e)(1-t)+p*(1-t) + m - (gr+e)]
Simplifying:
[(r+p)(1-t)+ m - gr -e*t)
Factors that influence
[(r+p)(1-t)+ m - gr -e*t)
Thus,
higher is r, p, or m, the higher is .
higher is gr, and e, the lower is .
Housing is often seen as a hedge against
inflation (i.e., the user cost of housing falls with
increases in e).
For rental housing, the maintenance
expenses are also tax deductible:
renter[(r+p)(1-t)+ m(1-t) - gr -e*t]
User Cost for Owners & Renters
owner[(r+p)(1-t)+ m - gr -e*t]
renter[(r+p)(1-t)+ m(1-t) - gr -e*t]
There are several reasons why we would not
expect the same user cost of capital for rental
vs. owner occupied housing
m higher for rental units ( renter
rental property riskier, thus higher r ( renter
m is tax deductible for owners of rental property
( rente
Another issue:
There is tax advantage to ownership in that the
imputed rent from home ownership is not taxed.
Which is higher? owner or renter?
• Use estimates of parameters.
• If t=0.28, mo=0.02, mr=0.03, ro=0.05, rr=0.06,
e=0.03, p=0.02 and gr=0.01
owner[(ro+p)(1-t)+ mo - gr -e*t]
owner
owner
renter[(rr+p)(1-t)+ mr (1-t) - gr -e*t]
renter
renter
If V=$100,000, we can
derive Rowner and Rrenter
Rowner=0.052*100000 =$5200 per year
Rrenter=0.0608*100000 =$6080 per
year
Thus, on same valued property, renters
pay about 17% more per year.
Distinguishing Features of
Renters and Owners
Low income typically rent. Why?
Down payment constraints.
Value of t typically lower for low income
households
High mobility households rent due to
the high transactions costs associated
with buying.
College students are highly mobile and
have low income. Almost universally rent.
Housing Demand
Which price (V or R) is correct?
QD=f(R, Ipermanent, demographic char’s)
Problem with Q?
Given multidimensional nature of Q, what is a
“unit” of housing?
Must somehow standardize. Techniques:
Isolate a class of housing
Use hedonic housing price technique to create
constant quality unit of housing (more later).
Empirical Findings on Demand
There are numerous housing demand studies.
General findings indicate:
Demand is slightly price inelastic to slightly elastic
(usually 0.7 - 1.2)
Primary determinant of price elasticity?
What does SUM suggest income effect would be?
Housing is income inelastic (approx. 0.75)
Income elasticity is lower for renters than owner occupants.
Income elasticity increases with income.
• 0.1-0.6 for low income, 0.7-1.1 for high income
The Hedonic Approach
Hedonic approach views housing as a
bundle of attributes.
As noted earlier, technique can be used to
derive measures of constant quality housing.
Derive hedonic price function.
P=f(Structural Chars, Neighborhood Chars, Fiscal)
Derive implicit prices for characteristics as
=P/Attribute.
Derive value of constant quality housing as the
sum of vector (*attribute).
Demand for Attributes
Can actually derive demand functions for
attributes using hedonic approach.
Overview:
Estimate implicit prices (=P/Attribute)
Fresno data can be used to derive these
In second stage, run the following regression
model: Attribute=f(,Income,Demographic)
Note, there is an interesting identification
problem.
Overview of findings:
Demand for Attributes: Findings
(Follain and Jimenez, 1985)
Demand for living space is income inelastic
(elasticity is about 0.5).
Demand for structural quality is income elastic
(ie., over 1.6)
Demand for neighborhood amenities is income
elastic.
Also, Muth finds demand for standard (as
opposed to substandard) housing is income
elastic.
Supply side of housing market
Little evidence of monopoly power.
Sources of supply include new and used
housing.
Housing deteriorates w/o maintenance,
and thus quality adjustments may lead to
a housing unit passing from one income
class to another.
Older housing deteriorates faster.
Housing Supply Function
Qs=f(V, Pinput,technology,expectations,govt)
Determinents:
Price is the asset price of housing.
Higher input prices reduce supply.
Technological advancements increase supply.
Expectations concerning future prices
influence supply.
(e.g., rent control can dampen supply)
Government regulation influences supply
e.g., zoning, min. square footage requirements,
Empirical Supply Functions
(Price elasticity<1.0)
Reasons:
New housing is a small fraction of the
overall stock of housing.
Thus, it cannot lead to large quantity
adjustments (for overall market) when price
changes.
Deterioration of used buildings into the next
lower quality classification.
Deterioration is slow.
Remodeling to upgrade quality.
Expensive.
Housing Market
Equilibrium
R-Q space: V-Q space:
(Map Supply curve) (Map Demand Curve)
R S V S
Re Ve
D D
Q Q
Qe Qe
What happens when increases?
(suppose real interest rates increase)
Rent-Quantity Space
# As user cost increases, S’
the supply curve is R
mapped to higher S
R’e
values of R for the
Re
same value of Q.
# This is because the
suppliers command D
higher rents when Q
their costs increase. Q’e Qe
What happens when increases?
(suppose real interest rates increase)
As user cost goes up, the Value-Quantity Space
demand curve is mapped
to lower values of V for
each Q. V S
This is because as rents Ve
increase, housing of a
given value is less V’e
desirable to consumers, D
and hence demand falls D’
at each Q. Q’e Qe Q
Role of Demographic
Characteristics on Housing
Prices
Mankiw and Weil published
a paper in late 1980’s
suggesting strong impact
from baby-boom.
Mankiw and Weil Briefly
Cross-sectional relationship (1970 Census)
between quantity of housing demanded
(ie., dollar value) and demographic
characteristics of household.
Specifically looked at age distribution of
population.
Findings: Demand rises sharply between age
of 20-30, and then declines after age 40 (by
about 1% per year).
What is impact of Baby
Boom?
Large cohort moving through population
distribution.
Findings suggest predictable influence of baby boom
on future values
They simulate impact, assuming housing
demand of those in 1970 is stable.
Findings suggest housing markets are not efficient.
What might be the problem with their methodology?
Housing Durability and Filtering
We now know something about the demand
& supply sides of housing market.
Furthermore, we know housing is durable.
This has several implications
First, as housing deteriorates, costs increase (due to
maintenance) and revenues decrease (due to
quality).
Second, housing may pass to different income class
over time (i.e., filter down)
Examine both phenomena.
Revenue and Cost
Adjustments
TC’
TR
TR,TC TR’
Suppose an
apartment building
gets older. TC
Demand falls as
quality declines.
Thus TR falls.
Costs go up as
maintenance
increases.
Thus, TC rises.
Q’ Q
Quantity falls
Eventually, property falls out
of this particular submarket.
If this is not the lowest
quality housing market,
housing will filter to other
users.
Filtering
Filtering:
A change over time in the position of a
given dwelling unit within the distribution
of housing rents and prices in the
community as a whole.
As a dwelling ages, it provides less
housing services per year, cet. par.
Materials deteriorate, and technology
becomes dated.
Look at Filtering Process
Assume all housing rented.
Assume constant population.
Consider 3 housing quality categories,
and 3 income groups.
Housing Quality: High, Medium, Low
Income Levels: High, Middle, Low
Assume an initial equilibrium.
Maximizing rents requires matching as
follows:
Equilibrium Sorting of
Households and Housing
Housing Quality
Low Medium High
Low X
Income Group Middle X
High X
Assume Income Growth of
Highest Income Group
Adjustment in Sorting
Housing Quality
Low Medium High Higher
Low X X
X X
Income Group Middle
High X X
Low quality falls
out of housing stock
Does this lead to permanently
higher housing quality for poor?
There is an equilibrium level of maintenance
for all income groups.
This will not increase as housing filters down.
Thus, improvement is temporary.
Depends on rate of depreciation.
If depreciation is slow, it can help.
If depreciation is rapid, it cannot.
This is an empirical issue.
Next time, we will look at the article by Weicher.
