the effects of rent stabilization and vacancy …...2010/04/19 · rent stabilization protects...
TRANSCRIPT
1
The Effects of Rent Stabilization and Vacancy Decontrol
on Rents, Rental Property Values and Rent Burdens
in Berkeley, California
April 19, 2010
Jay Kelekian, Executive Director
Stephen Barton, Ph.D., Project Manager
Berkeley Rent Stabilization Board
2025 Milvia Street
Berkeley, CA 94704
Tel: 510-981-RENT (7368)
E-mail: [email protected]
2
3
The Effects of Rent Stabilization and Vacancy Decontrol on Rents,
Rental Property Values and Rent Burdens in Berkeley, California
Contents Contents .......................................................................................................................................... 3 Summary ......................................................................................................................................... 4 Section 1: Overview of Rent Stabilization and the Rental Housing Stock ..................................... 6
The Rent Stabilization and Eviction for Good Cause Ordinance ........................................... 6 Rental Units Subject to Rent Stabilization.............................................................................. 7
Section 2. Berkeley Rents, 1978 to 2009 ...................................................................................... 11 Rent Levels in Berkeley, the Bay Area and the U.S. ............................................................ 11 Rent Increases Under Rent Stabilization in Berkeley, 1978 to 2009 .................................... 13
The Effects of Rent Stabilization on Tenancies Begun Prior to Vacancy Decontrol ........... 17
Effects of Rent Stabilization on Tenancies Begun After Vacancy Decontrol ...................... 18 The Overall Increase in Rents Resulting from Vacancy Decontrol ...................................... 21 Building Conditions and Maintenance .................................................................................. 21
Section 3: Berkeley Rental Property Values, 1978 to 2009.......................................................... 24 Sales Prices of Berkeley Rental Properties with Five or More Units ................................... 24 Comparison of Rental Property Sales Prices in Berkeley and Alameda County .................. 27
Trends in the Value of Multi-family Properties by Size ....................................................... 28 The Effect of Vacancy Decontrol on Property Values ......................................................... 29
The Effect of Vacancy Decontrol on City Tax Revenue ...................................................... 30 Price, Leverage and Risk ...................................................................................................... 31
Section 4: Fair Return and Low-Turnover Properties ................................................................... 34
Fair Return in Berkeley ......................................................................................................... 34
Effects of Tenant Turnover Rates on Rents .......................................................................... 37 Section 5: Changes in Rent Burden, 1980 – 2009 ....................................................................... 39
Bay Area Rent Burden, 1980 - 2008 ..................................................................................... 39
Berkeley Rent Burden, 1988 - 2009 ..................................................................................... 41 Section 6: The Economics of Rent Stabilization ......................................................................... 42
Section 7: Recommendations ........................................................................................................ 47 Appendix 1: Annual General Adjustments, 1981 - 2010 ............................................................. 49
Appendix 2: Numerical Tables for Figures .................................................................................. 50 Appendix 3: Vacancy Rate Data: Findings and Analysis ............................................................ 60 Appendix 4: Estimating Land Rent in the Bay Area Rental Housing Market .............................. 64
4
Summary
A majority of the housing units in Berkeley are rented and tenants make up half of Berkeley's
household population. Berkeley’s rent stabilization program has jurisdiction over three-quarters
of Berkeley rental units, those built prior to 1980 in properties with two or more units. The
program normally registers about 19,000 units each year, and potentially covers another 2,000
rental units that are temporarily exempt, most often because those units are occupied by tenants
who receive rent subsidy. Nearly 2,800 rental property owners own registered units, but more
than half are owned by the 281 owners with 15 units or more. Nearly 4,000 units are occupied by
tenant households who have been in place since 1998, before the beginning of state-mandated
vacancy decontrol. The remaining 15,000 units have tenants who moved in after 1998 and began
their tenancy at market rent.
Rents in the Bay Area have consistently increased faster than rents in the rest of the U.S. and are
higher than any other major metropolitan area. Current rents in Berkeley are consistent with this
trend. According to the American Community Survey the 2008 median monthly rent was $676
for all U.S. metropolitan areas, while Berkeley rent registration records showed a 2008 median
rent of $1,100. In contrast, for the one fifth of rent stabilized units that have never received a
vacancy increase since vacancy decontrol began in 1999 the 2008 median rent was $679.
Rent control delivers major economic benefits to the rent stabilized tenants who have not moved
since 1999 when vacancy decontrol went into effect and delivers modest benefits to later tenants,
whose initial rent is set at market but whose subsequent rent increases are limited to 65% of the
increase in the Consumer Price Index (CPI). As a result of vacancy decontrol, investors in rental
housing receive approximately $100,000,000 a year more in rent than they would have if
Berkeley had continued its strong rent control system and this amount increases as the remaining
long-term rent controlled tenants move out. Rent stabilization has reduced rents for long-term
tenants by $27,000,000 annually and for post-decontrol tenants by as much as $10,000,000.
The average sales price per unit for Berkeley rental properties with five units or more built prior
to 1980 has gone from $53,000 for properties sold in 1998, just prior to full vacancy decontrol, to
$162,500 in 2009 and the price per square foot increased from $78 in 1998 to $192 in 2009. The
value of the 19,000 pre-1980 rental units in Berkeley has increased by over $1 billion as a result
of vacancy decontrol. Just over 70% of Berkeley’s rental properties are still under the same
ownership as in 1998, so most owners have enjoyed major increases in property values.
Maintenance in Berkeley properties has improved somewhat since vacancy decontrol, but there a
still a substantial number of properties where tenants report poor conditions and multiple
physical problems with their buildings.
There are around 600 units in properties, mostly with one to four units, that had little or no tenant
turnover since 1998. The annual rent increase formula of 65% of the increase in the Consumer
Price Index (CPI), which replaced annual cost studies in 2005, may not provide a sufficient rent
increase in these properties. Past Rent Board studies found that in Berkeley average rent
increases for most properties needed to roughly match the increase in the CPI, although their can
be substantial variation in the rents of particular units. Where there is regular tenant turnover, the
vacancy increases make up the necessary difference. For the exceptional cases where there is
5
little or no turnover, the Rent Board should provide an expedited, low-cost procedure to allow
owners to request an increase in building rents to match the increase in the CPI.
In the Bay Area as a whole, rents have increased faster than both incomes and the cost of living,
and as a result the Bay Area’s low and moderate income tenants had less real income left after
paying rent in 2008 than tenants had in 1980. In post-vacancy decontrol Berkeley, rent takes an
increasing percentage of the income of Berkeley tenants. In 1998 41% of non-student tenants
spent more than Federal affordability standard of 30% of income for rent. This increased to 53%
in 2009 for non-student tenants who moved in after vacancy decontrol even though new tenants
have higher incomes. The rent burden did not change between 1998 and 2009 for the long-term
tenants whose rents were set prior to vacancy decontrol. Although its economic effects are
reduced, rent stabilization continues to protect tenants from displacement due to drastic increases
in rents such as occurred during the dot-com bubble.
The Bay Area housing market does not work the way markets are supposed to. Increases in rents
fail to generate sufficient additional housing production to hold rents down to levels that are
normal elsewhere in the U.S. As a result, Bay Area and Berkeley rents are based on scarcity and
the limits of what tenants can afford, rather than on the actual cost of operating and maintaining
the buildings. Rent stabilization protects tenants from rapid increases in rents while they remain
in the same apartment, but vacancy decontrol ensures that overall rents are at levels that are far
above the level necessary to actually operate and maintain rental housing. It remains an open
question what can be done to reduce the rent burden of low-income tenants.
6
Section 1: Overview of Rent Stabilization and the Rental Housing Stock
The Rent Stabilization and Eviction for Good Cause Ordinance
Berkeley Municipal Code Section 13.76: “Rent Stabilization and Eviction for Good Cause
Ordinance” has two main components, as described in its title.1 Rent stabilization limits rent
increases for most tenants who live in properties with two or more units constructed prior to
1980, which is the year the ordinance went into effect. Good cause for eviction requirements
apply to all rental units, even those units exempt from rent stabilization, with the exception of
duplexes that were owner-occupied in 1979 and are owner-occupied currently.2
Rent stabilization in Berkeley began in 1978, when the voters passed a local initiative that
temporarily reduced rents, requiring landlords to pass through to tenants the majority of the
operating cost savings the landlords received from the property tax reductions resulting from
passage of Proposition 13. During the campaign for Proposition 13, proponents had claimed that
reduced costs to owners would result in lower rents. Since rents continued to increase after its
passage, tenant advocates in a number of cities including Berkeley successfully proposed
measures to require rent reductions. Then, in 1980, Berkeley voters passed the Rent Stabilization
and Eviction for Good Cause Ordinance, making rent controls permanent, requiring rent
registration and establishing a Rent Stabilization Board to implement the requirements of the
ordinance. In 1982, the voters passed a Charter Amendment establishing an elected Rent
Stabilization Board (Berkeley Charter, Article XVII, section 121). From 1980 through 1998
rents in units built prior to 1980 were controlled permanently, so that the rent did not change
when a tenant moved out and new tenants moved in.3
In 1995 the State of California passed new legislation denying local governments the power to
place units under permanent rent controls. From 1996 to 1998 there was a phase-in period in
which owners could rent vacant units for 15% more than the previous rent ceiling. Then as of
January 1, 1999, local governments could no longer limit the initial rent for new tenancies,
although they can limit subsequent rent increases on the tenants as long as they remain in the
unit.4 This system is usually called “vacancy decontrol” although it is really “vacancy decontrol
– recontrol.” The new law also permanently removed all single-unit properties from rent control,
including both single-family houses and condominiums, except for tenancies that began prior to
1999.
1 Berkeley Municipal Code, Chapter 13.76, et seq.
2 Eviction for good cause does not apply when a tenant does not rent a unit but rather shares one with the owner.
3 For a detailed history of Berkeley’s rent stabilization program and changes in the ordinance up to 1994, see
Planning & Development Department, City of Berkeley, Rent Control in the City of Berkeley: 1978 to 1994: A
Background Report, May 27, 1998. 4 California Civil Code sections 1954.50 through 1954.535
7
In 2004 voters approved an amendment to the Rent Stabilization Ordinance to cease doing
annual cost studies and simply index the Annual General Adjustment to the rent at 65% of the
increase in the CPI-All Items.5
The major purpose of the Rent Stabilization and Eviction for Good Cause Ordinance is to
provide a stable housing environment for tenants while assuring that landlords are able to receive
a fair return on their investment. It assures tenants in rent-stabilized units that once they move in
their rents will not drastically increase, giving tenants a level of security similar to that of
homeowners who are protected from rapid cost increases by the state property tax limitation and
fixed-rate mortgages. In addition to annual rent increases (through the Annual General
Adjustment, or AGA), landlords can apply for individual rent adjustments if the increases they
receive through tenant turnover and the AGA are not sufficient to provide them with the legally
required fair rate of return. The ordinance also protects tenants from arbitrary evictions through a
system of eviction controls and twelve defined just causes for eviction. Good cause for eviction
requirements apply to virtually all rental units, including those built after 1980, condominiums
and single-family houses.
Rental Units Subject to Rent Stabilization
Berkeley currently has approximately 27,700 rental units.6 Virtually all of these rental units are
subject to Berkeley’s “good cause for eviction” ordinance. About 21,000 of them - about 75% of
Berkeley’s rental housing stock - are currently or potentially subject to rent stabilization.
The major categories of units not subject to rent stabilization are:
Units built after 1980. There are approximately 2,700 such rental units.
Single-family or condominium units. Currently there are only 144 tenants in one-unit
properties who moved in prior to 1999 and are still covered by rent stabilization.7 There
are approximately 4,000 exempt single unit rentals.
Rental units in two-unit properties which were owner-occupied on December 31, 1979 as
long as an owner occupies the other unit. These units are covered by the ordinance when
an owner does not occupy the other unit. Currently there are 270 such exempt units.8
5 The Rent Board and the Berkeley Property Owners Association jointly proposed this amendment in settlement of
litigation brought by the BPOA challenging the Rent Board’s practice of reducing increases for maintenance of net
operating income by the amount of past increases that were more than legally necessary. 6 The U.S. Census found that as of April 1, 2000 Berkeley had 46,875 housing units, of which 26,498 were either
rented or vacant and for rent. By the end of 2009 approximately 1,200 additional multifamily rental units were
completed, for a total of about 48,100 units and 27,700 rentals. The Census counts rented rooms as separate units if
they have direct access to public spaces, such as a hallway, leading to an exit. Otherwise they are considered group
quarters. According to the Housing Department, the Rental Housing Safety Program (RHSP) inventory includes
25,708 rental units and 3,011 rooms for rent. This inventory likely includes many rooms that the Census counts as
group quarters but likely also misses a number of rented single-family homes, which are often rented only
temporarily. 7 As of August 3, 2009.
8
Units whose tenants receive rental assistance from the government through programs
such as the Section 8 Housing Choice Voucher Program or Shelter Plus Care are not
covered by rent stabilization as long as the landlord does not charge a rent above the
“payment standard” set by the Berkeley Housing Authority. Currently there are 1,426
such units. When occupied by a tenant who is not receiving rental assistance, these units
are subject to rent stabilization unless another reason for exemption applies.
Units in non-profit housing provided at below-market rents to low-income tenants are
exempt unless the occupant is an original tenant who moved in prior to the non-profit
organization taking ownership. Currently there are 114 such units.
A total of about 21,000 rental units are normally subject to rent stabilization.9 For the 2008-2009
fiscal year landlords paid registration fees for 19,050 units, indicating that they were subject to
rent stabilization. Another 1,810 units were reported exempt, either because they were occupied
by tenants receiving rental assistance or because they were in a 1979 owner-occupied duplex.
There are also as many as 1,000 units in multifamily properties that would be subject to rent
stabilization if they were rented out, but are currently owner-occupied or otherwise used by the
owner, occupied rent free, or vacant and not available for rent.10
This suggests that the maximum
potential coverage of Berkeley’s rent stabilization ordinance is around 22,000 units.
Physical Characteristics of the Rental Housing Stock
Berkeley’s rental housing stock is quite diverse, with about one-quarter in one- and two-unit
buildings, one-sixth in each of the following size categories: 3-4 unit buildings, 5-9 unit
buildings and 10-19 unit buildings, and one quarter in buildings with 20 or more units. Almost
all of the rental housing constructed since 2000 has been in buildings with 20 units or more.
Table 1 compares the 2000 Census data for rental units by the number of units in the building
with 2009 Rent Board data for rent-stabilized units and units temporarily exempt while occupied
by tenants receiving monthly rental assistance, mostly through the Section 8 Housing Choice
Voucher program. There are differences of definition to deal with, for example two duplexes on
8 As of August 3, 2009.
9 The registered and exempt units almost exactly match the 2000 Census data. According to the 2000 Census,
24,028 of Berkeley’s 25,748 occupied rental units were built prior to 1980. The Census also reports 750 units that
were vacant and for rent, but does not give a construction date for them. Assuming these are similar to the occupied
units, we can add another 700 rental units to the total built prior to 1980. Of this total of 24,728 rental units built
prior to 1980, the Census reports that 4,033 of these were in single-family houses, and all but 144 of these are now
permanently exempt due to tenant turnover since 1999. This leaves about 20,850 units subject to rent stabilization.
There are two offsetting sources of error in this estimate. The Census counts units in the structure, rather than units
on the property. Berkeley has a number of multi-unit properties with separate cottages, so single-family rental units
of this type are generally subject to rent control. In addition, some rental units in pre-1980 multi-family buildings are
not subject to rent control because they are condominiums. 10
According to the 2000 Census there are a total of 25,021 units in multifamily structures in Berkeley, of which
about 23,400 units were built prior to 1980. Of these, about 2,300 units were owner-occupied at the time of the 2000
Census. At least 1,000 of these owner-occupied units are condominiums and thus permanently exempt from rent
stabilization. Another 270 are in 1979 owner-occupied duplexes that are exempt as long as they are owner-occupied.
9
one lot would count as two duplex rentals for the Census and one four-unit rental for the Rent
Board, but the broad trends are clear. Exempt rentals are concentrated in single-family units,
smaller multi-family properties where condominium conversion has been concentrated, and in
newly constructed larger properties.
Table 1: Renter-occupied units by units in structure
All 2000 Rentals by 2009 Registered Rentals by
Units in Structure Units in Property
1 unit 4,428 152
2 units 2,504 1,201
3 or 4 units 4,282 3,299
5 to 9 units 4,425 4,491
10 to 19 units 4,258 4,082
20 to 49 units 4,163 4,340
50 or more units 1,666 1,163
Total units 25,748 18,728
Sources for Table 1: U.S. Census Bureau, 2000 Census (does not include rentals built after April 1, 2000, all of
which are exempt from rent stabilization); Berkeley Rent Stabilization Program, August 25, 2009 (point in time data
gives several hundred fewer units than the total for registrations during a full year.) The 2000 Census is used
instead of the 2006-8 American Community Survey because there are major discrepancies between the Census and
ACS unit totals and the Census count methodology is more accurate than the ACS survey.
Table 2 shows the size of Berkeley’s rental housing by number of bedrooms. Exempt units are
fairly evenly distributed among zero, one and two-bedroom units, but make up more than half of
rental units with three or more bedrooms because many of the larger rental units are in single-
family houses that are exempt.
Table 2: Renter-occupied units by number of bedrooms
All Rentals Registered Rentals
Bedrooms in Unit
No bedroom 5,184 3,792
1 bedroom 10,532 7,890
2 bedrooms 7,371 5,677
3 bedrooms 1,945 873
4 or more bedrooms 716 293
Total units 25,748 18,525
Sources for Table 2: U.S. Census Bureau, 2000 Census (does not include rentals built after April 1, 2000); Berkeley
Rent Stabilization Program, August 25, 2009 . Point in time data gives several hundred fewer units than the total for
registrations during a full year. Several hundred units with missing data are excluded. Units with no bedrooms
include both studio apartments and rooms in a rooming house or residential hotel as long as the rooms have
independent access to hallways and entry areas.
10
The distribution of ownership of registered rental units in Berkeley is very diverse, ranging from
a family that owns only one unit to a family that owns over a thousand units. Table 3 shows the
distribution of ownership by number of units owned as reported to the Rent Stabilization Board.
This table somewhat overstates ownership dispersal, because the Rent Board data do not take
into account situations in which one owner or a group of owners use different corporations or
partnerships. Two-thirds of the owners have from one to four registered units, but all together
they own only 22% of all the registered units. The ten percent of owners with 15 units or more
own over 50% of the registered units. The average number of registered units per owner is 6.7.
Table 3: Ownership by Number of Registered Rental Units11
Separate Owners Currently Registered Units
Total 2,792 18,732
Owners with 1-4 units 1,876 4,104
Owners with 5-14 units 635 5,170
Owners with 15-29 units 193 3,996
Owners with 30+ units 88 5,462
Owners with 1-14 units 2,511 9,274
Owners with 15+ units 281 9,458
11
The number of units owned does not include rental units exempt from registration, such as units rented to
tenants with Section 8 vouchers.
11
Section 2. Berkeley Rents, 1978 to 2009
Rent Levels in Berkeley, the Bay Area and the U.S.
Rents in Berkeley, as in the Bay Area as a whole, are among the highest in the United States. The
major exception is units that have not received a vacancy decontrol increase.
Table 4: 2008 Contract Rent in 25 Largest Metropolitan Areas and Berkeley Area
Lower
quartile Median
Upper
quartile
San Jose-San Francisco-Oakland, CA CSA $882 $1,188 $1,623
Washington DC-Arlington, VA-MD-WV $850 $1,127 $1,484
San Diego-Carlsbad-San Marcos, CA $861 $1,126 $1,524
Alameda County, CA $828 $1,091 $1,438
Berkeley, CA $762 $1,069 $1,488
Los Angeles-Long Beach-Santa Ana, CA $790 $1,046 $1,420
Boston-Cambridge-Quincy, MA-NH $653 $983 $1,323
New York-N. New Jersey-Long Is., NY-NJ-PA $678 $956 $1,295
Riverside-San Bernardino-Ontario, CA $709 $950 $1,247
Miami-Fort Lauderdale-Pompano Beach, FL $702 $923 $1,219
Oakland, CA $694 $912 $1,204
Sacramento--Arden-Arcade--Roseville, CA $707 $892 $1,184
Seattle-Tacoma-Bellevue, WA $656 $839 $1,125
Baltimore-Towson, MD $601 $812 $1,094
Phoenix-Mesa-Scottsdale, AZ $607 $777 $1,015
Chicago-Naperville-Joliet, IL-IN-WI $602 $766 $978
Minneapolis-St. Paul-Bloomington, MN-WI $606 $761 $969
Philadelphia-Camden-Wilm., PA-NJ-DE-MD $561 $748 $960
Denver-Aurora, CO $578 $743 $979
Tampa-St. Petersburg-Clearwater, FL $588 $741 $948
Atlanta-Sandy Springs-Marietta, GA $588 $738 $905
Portland-Vancouver-Beaverton, OR-WA $596 $720 $913
Berkeley, CA -- pre-1999 tenancies $599 $679 $813
United States $466 $676 $965
Dallas-Fort Worth-Arlington, TX $531 $665 $848
Houston-Sugar Land-Baytown, TX $507 $638 $827
Detroit-Warren-Livonia, MI $497 $637 $810
Cincinnati-Middletown, OH-KY-IN $423 $549 $708
Pittsburgh, PA $362 $495 $643 Table 4 sources: U.S. Census Bureau American Community Survey 2006-2008 (ACS). Berkeley 2009 Tenant
Survey median for pre-1999 tenancies was $697, less the 2.7% AGA increase is $676. The U.S. median is at the
low end because it includes all U.S. metropolitan areas, not just the 25 largest shown in this table.
12
The median Berkeley rent is slightly lower than that of Alameda County, but Berkeley rental
units are significantly smaller, with a median of 3.4 rooms each compared with 3.8 for Alameda
County and the Bay Area as a whole.
Figure 1 shows the long-term trend of San Francisco Bay Area and U.S. real rents.12
The term
“real” in this context means adjusted for inflation.13
When the other items people buy increase in
price faster than rents, then “real” rents go down and when rents increase faster than the other
items people buy, “real” rents go up.
Prior to 1959, changes in Bay Area rents closely tracked changes in rents nationally. Since then
Bay Area rents have increased while nationwide rents have remained largely stable. The Bay
Area real rent has also been more volatile, including the rapid rise and decline caused by the dot-
com bubble. Figure 1 shows that by 2008, Bay Area real rents had increased by nearly 170%
12
“Bay Area” statistics from the U.S. government are based on the San Jose-San Francisco-Oakland Consolidated
Statistical Area covering the eleven counties of Alameda, Contra Costa, Marin, Napa, San Francisco, San Benito,
San Mateo, Santa Clara, Santa Cruz, Solano and Sonoma. The nine-county area covered by the Association of Bay
Area Governments does not include Santa Cruz and San Benito Counties. 13
The CPI-Rent is adjusted by the CPI-All Items Less Shelter because increases in rents are included in the overall
CPI, with housing costs making up one-third of the total index. Using the CPI-All Items would place rents in both
the numerator and denominator of the adjustment and this circularity would minimize the actual increase. The
change in U.S. real rents is adjusted by the U.S. CPI-All Items Less Shelter and the Bay Area CPI-Rent is adjusted
by first the U.S. CPI-LS from 1935-76 and then the Bay Area CPI-LS for 1976-2008 because the Bureau of Labor
Statistics only provides the CPI-LS for the Bay Area since 1976 although it provides it nationally since 1935.
13
over U.S. real rents, which is the main reason why the median Bay Area rent is 175% of the
median U.S. rent as shown in Table 4.
Rent Increases Under Rent Stabilization in Berkeley, 1978 to 2009
Figure 2 shows median Berkeley rents from 1978 to 2009. The line that starts in 1978 shows the
median rents under Berkeley’s rent stabilization system.14
The AGA figures in this and all
subsequent figures and tables underestimate the actual increases between 1980 and vacancy
decontrol in 1999 and illustrates the most restrictive possible situation. Diverging off from the
AGA line in 1996 is a line showing the rate of increase in the average rent of all controlled units,
including those that received a vacancy increase. Hovering above both lines is a line showing the
median vacancy rent for a one-bedroom apartment.15
14
This figure assumes that all increases after 1979 were only those allowed under the AGA, so all individual rent
adjustments (IRAs) are omitted, and it assumes that there was partial compliance between 1978 and 1980 prior to
rent registration so that rents on average were flat rather than reduced between 1978 and 1979 and increased by 5%
in 1980. The starting point is the median Berkeley rent of $223 reported in the 1980 Census. In some years the AGA
was a fixed dollar amount, so that units with rents below the median had a higher percentage increase and units with
rents above the median had a lower percentage increase. 15
Medians for all vacancy registrations are not available, but the one-bedroom rent is usually only slightly below
the median rent for all units since over half of Berkeley’s registered rental units are one-bedroom or less.
14
The AGA line does not include the additional “utility allowance” increases for those units, about
a quarter of the total, in which the landlord pays for gas or electricity within the individual unit.
It also leaves out individual rent adjustments made for over 6,000 units to compensate owners for
“historically low rents”, capital improvements and various other purposes.16
The year 1991 is
shown twice in order to show the special increase for maintenance of net operating income that
was added to the normal AGA in that year pursuant to a court order.
Within the history of rent stabilization in Berkeley there are distinct periods with quite different
policies. From 1978 to 1980 there were a series of temporary ordinances. In 1978 voters passed
an initiative under which rents were supposed to be rolled back by 80% of the amount of the
decrease in property taxes, estimated to be approximately 7% of the rent for an average unit, with
increases allowed only for operating costs over and above the landlords’ 20% of the decrease in
property taxes. This was followed by another temporary rent ordinance which allowed a limited
rent increase of up to 5% in 1980 to cover increased operating expenses.17
At that time there was
no agency with the administrative capacity to systematically enforce the requirements of the
ordinances. Surveys conducted in 1979 and 1981 found widespread non-compliance and an
informal practice of vacancy decontrol during this period.18
Figure 3 follows this finding and
assumes partial compliance during this period averaging out to a rent freeze from 1978 to 1979,
rather than a roll-back of rents as required under the ordinance.
In 1980 a voter initiative established the Rent Board and a process for setting general and
individual rent increases and required landlords to register all rents. From 1980 to 1991 the
Annual General Adjustments (AGAs) increased by 61%, compared with the increase in the Bay
Area CPI of 72%. AGAs were based on an annual operating cost study and no increase was
provided for Net Operating Income (NOI). NOI increases were allowed at 40% of the rate of
inflation through an individual adjustment process. In 1990 the California courts ruled that this
did not provide owners with a fair rate of return because it was overly burdensome to expect
each owner to separately file for an individual rent adjustment to gain an increase for NOI and
the 40% of CPI formula was not adequately supported.
In 1990 the courts mandated that the Board provide an increase in AGA for maintenance of NOI
retroactive to 1979 in order to meet constitutional fair return standards. The year 1991 is entered
twice in Figure 3 in order to show the major increase that resulted. The Rent Board decided to
provide a retroactive NOI increase indexed at 100% of the CPI-Less Shelter going back to 1979.
The City Council sued on the grounds that the increase was excessive, presenting evidence that
NOI increases could reasonably be indexed at 50% of the increase in the CPI. This could provide
rental property owners with a fair return because most investors in rental property leverage their
investment by borrowing a substantial part of the purchase price of the property. For example, if
the owner’s mortgage payments take up half of the NOI, leaving the other half for cash flow,
16
Planning & Development Department, City of Berkeley, Rent Control in the City of Berkeley: 1978 to 1994: A
Background Report, May 27, 1998, p.148 indicates 6,326 IRAs made by 1993. 17
For a detailed discussion of the effect of Berkeley’s temporary rent ordinances see “The Temporary Ordinances:
June 1978 – May 1980”, pp. 139-142 in Planning & Development Department, City of Berkeley, Rent Control in the
City of Berkeley: 1978 to 1994: A Background Report, May 27, 1998, Appendix A. 18
Ibid.
15
then an increase in NOI at 50% of the rate of increase in the CPI would increase the landlord’s
cash flow sufficiently to maintain its real, inflation-adjusted value.19
The courts ruled that the Board had substantial flexibility in choosing an appropriate percentage
increase and held that while the increase granted might or might not be good public policy, it was
within the discretion of the Board. After changes in its composition, the Board it changed its
regulations. A 1995 study determined that indexing NOI at 65% of the CPI-Less Shelter would
provide a fair return and meet constitutional requirements and the Board adopted this standard.20
From 1996 to 2005 the Board held annual AGA increases to the amount necessary to cover
increases in operating expenses and provided little or no increase for maintenance of NOI on the
grounds that previous increases had surpassed what was necessary under the new standard. By
2005 this process of absorbing the “excess” NOI increase was completed.21
In 2004, voters approved an amendment to the Rent Stabilization Ordinance to cease the annual
cost studies and simply index the Annual General Adjustment to the rent at 65% of the increase
in the CPI-All Items, beginning with the 2005 AGA. The Rent Board and the Berkeley Property
Owners Association jointly proposed this amendment in settlement of litigation brought by the
BPOA in opposition to the Board’s AGA policy.
Vacancy decontrol began in 1999 just as the dot-com bubble took off in the Bay Area. Market
rents in Berkeley and throughout the Bay Area increased dramatically, reaching a peak in 2001.
Rents declined somewhat between 2001 and 2004, then began to rise again. As shown in Figure
3, median Berkeley rents continued to increase as long-term tenants gradually moved out and
new tenants moved in at market rate. Despite the recession of 2008-2009, Berkeley vacancy rents
in the first half of 2009 were virtually unchanged from the first half of 2008 and then declined by
3% in the third quarter of 2009 compared with the third quarter of 2008. 22
Bay Area rents
stabilized in 2009 at a level 3% higher than the average for 2008.23
Figure 3 compares the rate of increase in Berkeley rents with the increase in the Bay Area
Consumer Price Index since 1978. Each indicator starts at 100 in 1978 so that all increases refer
back to that starting point.
19
The clearest and most comprehensive analysis of fair return standards is Kenneth K. Baar, “Guidelines for
Drafting Rent Control Laws: Lessons of a Decade”, Rutgers Law Review, V.35,No.4, Summer 1983, pp.723-885. 20
Kenneth K. Baar, “Issues and Options for Rent Increase Standards Under Berkeley’s Rent Stabilization
Ordinance”, October 26, 1995. 21
Memorandum dated October 18, 2004 from Jay Kelekian, Executive Director, to Members of the Rent
Stabilization Board, “2005 Annual General Adjustment”, pp.4-5. 22
Berkeley Rent Board, “Market Medians: January 1999 through June 2009”, September 18, 2009. 23
Bureau of Labor Statistics, Consumer Price Index – All Urban Consumers, Rent of Primary Residence. San
Francisco-Oakland-San Jose.
16
Figure 3 shows that during the strong rent control period, and for long-term tenancies where an
owner has never received a vacancy increase, Berkeley’s rent control system has usually held
rent increases to somewhat below the increase in the consumer price index. (Berkeley’s AGAs
increased rents at roughly the same rate as the CPI-Less Shelter.) After vacancy decontrol,
however, overall rents in units subject to rent stabilization increased far above the cumulative
increases in the CPI. The current median Berkeley rent is 155% of what it would have been if
rents had been indexed to the CPI. In other words, more than one third of the current median rent
is due to vacancy decontrol.
Figure 4 shows that Berkeley rents have increased even more than Bay Area rents over all.
During the period from 1978 to 2009, median registered rents in Berkeley have increased by
438% while Bay Area rents increased by 377% as measured by the residential rent component of
the Consumer Price Index.
17
The Effects of Rent Stabilization on Tenancies Begun Prior to Vacancy Decontrol
The economic effects of rent stabilization vary depending on when the current tenancy began.
They are most substantial for units with long-term tenants that have been in place since 1998
because that means the unit’s rent has been under continuous control since 1980.24
The total
number of registered units with tenancies that began prior to January 1, 1999 is around 4,000,
which is 21% of all registered units and 15% of all rental units in the City.25
In order to estimate the total rent discount provided to long-term tenants we can compare the
mean rent of $745 reported in the 2009 tenant survey for pre-vacancy decontrol tenancies with
the mean rent of $1,301 that the survey reported for post-vacancy decontrol non-student
24
The figures for long-term tenancies include an unknown but very small number of units with more recent
tenancies where the landlord was not allowed to charge a market rent because of the circumstances under which the
former tenant left the unit. 25
Currently there are about 4,000 units that have not registered a vacancy increase with the Rent Board since
1999 when vacancy decontrol began. In the Rent Board’s 2009 tenant survey 15% of the units where the Rent Board
had no record of a vacancy rent increase the tenant reported that they had moved in after January 1999. In some
cases the landlord may have been denied a vacancy increase due to the circumstances under which the previous
tenant left the unit. There may also be some misclassified units, so the total may be anywhere from 3,600 to 4,200.
18
tenancies. This removes the student tenancies, which tend to be located in different areas and
have higher rents. Applying this difference of $556 monthly, $6,672 annually, to 4,000 units
results in an estimated total discount of $26,700,000.
Effects of Rent Stabilization on Tenancies Begun After Vacancy Decontrol
The economic effects of rent stabilization on any particular housing unit in Berkeley largely
depend on when the current tenancy began. Table 5 shows the rate at which units turned over
after full vacancy decontrol began on January 1, 1999.
Table 5: First-Time Vacancy Registrations, 1999 to 2008
Year
1st Time
VRs
Units Turning Over
with Pre-1999 Tenants
As % of All Units
Total Units
with VRs
% of All
Units With
No VR
Registered
Units
1999 3,668 19.1% 3,668 80.9% 19,169
2000 3,226 16.7% 6,894 64.3% 19,290
2001 1,549 8.1% 8,443 56.1% 19,236
2002 1,613 8.5% 10,056 47.2% 19,029
2003 1,246 6.6% 11,302 39.7% 18,758
2004 861 4.6% 12,163 34.8% 18,652
2005 630 3.4% 12,793 30.5% 18,418
2006 617 3.3% 13,410 27.6% 18,534
2007 494 2.7% 13,904 25.0% 18,545
2008 610 3.2% 14,514 22.8% 18,798 Source: Berkeley Rent Stabilization Board. Note that the total registered units each year is a point in time figure
which varies depending on how many owners miss registration payments each year.
The simultaneous arrival in 1999 of vacancy decontrol and the dot.com boom created a
particularly dramatic difference between the rents of tenants who moved in before and after
January 1, 1999. When the dot-com boom collapsed, Bay Area rents stabilized for several years,
and Berkeley rents have only recently risen above the 2001 level. As a result, when people
discuss rent stabilization in Berkeley, they often refer to the units occupied by pre-1999 tenants
as “rent-controlled units,” and describe more recent tenancies as “market rate units,” lumping
units that have received a vacancy increase since 1999 with exempt units. This shorthand way of
describing the Berkeley rental market obscures the more complex reality and the benefits that
rent stabilization provides to more recent tenants.
Figure 5 illustrates the economic effects of rent stabilization depending on when a tenancy
began. One solid line shows the rent ceiling of a long-term tenant who has been in a one-
bedroom apartment since 1998, prior to vacancy decontrol, starting with the median one-
bedroom rent in 1998. Another solid line shows the median vacancy rent reported to the Rent
Board for newly rented one-bedroom units starting in 1999 when vacancy decontrol began.
Dotted lines show the rent ceilings for these one-bedroom units in subsequent years under rent
stabilization and how they vary depending on the year the tenant moved in. (Under Berkeley’s
19
current rent stabilization system, after a tenant moves in no rent increase is allowed for the first
full calendar year after move-in, so that if a tenancy began in June 2006, the first AGA was
allowed on January 1, 2008 rather than 2007.)
Figure 5 illustrates the modest but real economic benefits that rent stabilization provides to the
majority of tenants who moved in at market rents after vacancy decontrol but who still receive
protection from rising rents. The 2009 Tenant Survey found that about 43% of all tenancies
began between 1999 and 2007, and most of these have directly benefited financially from rent
stabilization to some degree. Tenancies that began in 1999 were shielded from rising rents and
displacement caused by the dot-com bubble, and tenancies that began after the bubble collapsed
have been shielded from subsequent increases. Only those who moved in at the top of the market
in 2001 have as yet received no economic benefit from rent stabilization.26
The most recent
tenants, the approximately 38% who moved in during 2008 and 2009, have not been in place
long enough for rent stabilization to have a direct financial effect, but they benefit from good
cause for eviction and the knowledge that their future rents will remain stable.
26
When market rents decline, a tenant may ask for and receive a rent reduction from the landlord but this does not
change the rent ceiling, which is based on the actual rent charged during the first year. In subsequent years the rent
can be raised to the ceiling at any time with 30 or 60 days advance notice to the tenant.
20
The pattern shown in Figure 5 for tenancies since 1999 is characteristic of a rent stabilization
system. Tenants are protected against displacement by rapid rent increases and, as with
homeowners, are rewarded economically for remaining in one place. The stability and security
provided by the combination of rent stabilization and good cause for eviction is also very
important to many tenants, but this is not something that is readily given a dollar value. For the
community, residential stability is generally correlated with civic involvement, such as
participation in neighborhood watch and disaster preparedness groups and voting. Encouraging
such stability may be all the more important in a city with a large student population that results
in unusually rapid turnover in many neighborhoods, but again such benefits are impossible to
quantify with dollar values.
Table 6 shows the difference between the median 2009 vacancy rent for a one-bedroom
apartment and the 2009 rent ceiling for a tenant who moved in under each year’s median vacancy
rents from 1999 to 2008.27
The difference in monthly rent represents the potential savings that
rent stabilization provides tenants. The difference ranges from 0% to 14% and averages 5% of
the 2009 vacancy rent, an average of $800 annually. The difference for tenants in two-bedroom
units ranges from 0% to 16% and averages 7% of the 2008 vacancy rent, an average of $1,400
annually. Applying the move-in dates reported from the 2009 tenant survey, we can project that
the total economic benefit to them from rent stabilization is approximately $10.5 million a year
even under vacancy decontrol. We cannot say that the entire tenure discount is created by rent
stabilization, however. Even without rent regulation many landlords hold down increases for
tenants who remain in place, since changing tenancies creates costs and uncertainties.28
Table 6: 2009 Rent Stabilization Savings to Tenants
(Median One-Bedroom Unit)
Move-in Year Move-in (Market) Rent
2009 Rent Ceiling (Move-in Rent plus AGAs)
Savings over 2009 Market Rent
1999 $950 $1,097 14%
2000 $1,100 $1,252 2%
2001 $1,200 $1,329 0%
2002 $1,150 $1,274 0%
2003 $1,100 $1,204 6%
2004 $1,050 $1,139 11%
2005 $1,095 $1,179 8%
2006 $1,100 $1,155 9%
2007 $1,200 $1,232 3%
2008 $1,275 $1,275 0%
2009 $1,275 $1,275 0%
27
The 2009 rent is for the January through September period. Rent Stabilization Board, “Market Medians:
January 1999 through September 2009”, December 11, 2009. 28
W.A.V.Clark and Allan D. Heskin, “The Impact of Rent Control on Tenure Discounts and Residential
Mobility”, Land Economics, 58:1 (February 1982) 109-117. Clark and Heskin found rent discounts for tenants in
Los Angeles prior to passage of rent stabilization in 1979, with the discount becoming substantially higher under
rent stabilization, particularly for long-term tenants.
21
The Overall Increase in Rents Resulting from Vacancy Decontrol
The cost of vacancy decontrol to tenants and its benefit to landlords is currently about $100
million annually and this amount increases each year, as more units receive vacancy increases
and as rents increase faster than vacancy rent regulation would have allowed.
This estimate is based on a comparison of current rents with those that would have been allowed
under a strong rent control system in which rents increased at the same rate as the increase in the
consumer price index since 1978. In 1980 the Census reported the mean Berkeley contract rent
was $245. Based on Figure 2 and the associated table in the appendix, we estimate that the 1978
rent was 5% lower. Indexing this rent to the CPI starting in 1978 the current mean rent would be
$808. Comparing that to the current mean contract rent of $1,252 yields a differential of $444
monthly and $5,328 annually that applies to 19,000 registered rental units. This results in an
estimate of $101 million in increased rent annually, with the current annual rent amounting to
$285 million annually and the hypothetical annual rent under continued strong rent control
amounting to $184 million.
Table 7 compares current median market rents with median rents as they would have been
without vacancy decontrol.29
This is likely somewhat of an overestimate, since as previously
mentioned vacancy rents are skewed towards higher rent areas.
Table 7: Median 2009 Rents for Tenancies Starting Prior to 1999 and in 2009
Median 2009 Rent With Jan.-Sept. 2009
Unit size No Vacancy Decontrol Median Vacancy Rent Difference
Studio $657 $975 $318
1 BR $774 $1,275 $501
2 BR $959 $1,765 $806
3 BR $1,317 $2,450 $1,133
All Units $797 NA NA
Building Conditions and Maintenance
Some of the increased revenue from vacancy decontrol is clearly being spent on improved
maintenance. However there are still a substantial number of properties where tenants report
unresponsive owners or managers and units in poor condition, despite major rent increases.
29
1998 median rents are updated with subsequent AGAs to 2004 followed by a 100% of CPI adjustment instead of
the 65% of CPI adjustment formula because without vacancy decontrol AGAs would have continued to approximate
the increase in CPI. This increases the hypothetical controlled rent by 3.8% over the rent that results from the
current formula.
22
In the 2009 tenant survey, 75% of respondents reported that there is a physical problem in their
building. The most frequently listed were “doors or windows” at 38%, “plumbing” at 30% and
“mold” at 26% and “heat” at 18%. In addition, 25% reported problems with “noise or other
tenants”. This is a very substantial set of building problems. Nonetheless, it is slightly better
than in 1998, when 83% reported having one or more problems in their building. In addition, the
number of problems in each building has declined from an average of 3.5 per building with
problems to 2.4 per building with problems.
Table 8: Problems Reported in Building, 2009 and 1998
2009 1998 Decrease
Doors/Windows 38% 45% -7%
Plumbing 30% 37% -7%
Mold/Mildew 26% 46% -20%
Paint 19% 38% -19%
Heat 18% 27% -9%
Security/Lighting 16% 36% -20%
Secure Mailboxes 16% NA NA
Appliances 15% 24% -9%
Roof 9% 20% -11%
Stairs/Porch 9% 16% -5%
Elevator 6% NA NA
Since three-quarters of the respondents reported physical problems in their building, it is not
surprising that 76% of tenants have complained to the landlord or building manager at some time
during the previous year. Among those with complaints, 70% reported that the owner or manager
responded quickly, 26% said that they responded after repeated complaints, and 12% reported
that there was a complaint that the owner or manager did not respond to. Two-thirds (65%)
reported getting a problem fixed in less than 30 days and another 12% got the problem fixed in
more than 30 days. One third reported a problem either only partially fixed or not fixed,
including 19% reporting a problem was not fixed. (These figures add up to more than 100%
because respondents could pick more than one type of response and more than one type of
outcome.) The proportion describing their unit as being in poor condition (8%) is not much
changed from 9% in 1998 and 11% in 1988.
The substantial number of buildings with unresolved problems cannot be explained by the
remaining pre-vacancy decontrol tenancies. The 2009 tenant survey did not find clear differences
in reported unit conditions and building problems between pre- and post-vacancy decontrol
tenants. Pre-vacancy decontrol tenants are somewhat more likely to describe their unit as being
in “fair” rather than “good” condition, but they are somewhat less likely to report physical
problems with their building. The only substantial difference between the two groups is that 30%
of pre-vacancy decontrol tenants report problems with paint compared with 17% of post-vacancy
decontrol tenants.
23
24
Section 3: Berkeley Rental Property Values, 1978 to 2009
Sales Prices of Berkeley Rental Properties with Five or More Units
Rental property values are closely tied to rent levels. Since smaller multifamily properties with
2-4 units can also be in the homeownership market, we will focus on properties with five units or
more. Figure 6 shows the trend in Berkeley’s rental property values for the 32 years from 1977 to
2009.30
The solid line shows the average sales price per unit, and the dotted line shows how the
1978 price would have increased if it went up with the consumer price index, meaning that the
“real” inflation-adjusted price remained the same. There are two distinct periods, first the strong
rent control period from 1980 to 1998 in which property values increased at roughly the same
rate as the consumer price index, going from $16,150 to $52,950 per unit, and second, the post-
vacancy decontrol period from 1999 to 2009 in which values have increased more rapidly, from
$52,950 to $162,500 per unit.
30
Alameda County property records, accessed through RealQuest. Data from this source is extremely limited or
non-existent prior to 1977. Transactions were removed when they involved properties built since 1980, sales of
fractional interests, or where the property records hold internally contradictory data. The number and mix of
transactions varies from year to year. During the recession of 1981-82 there were very few property sales as well as
reduced prices.
25
Figure 7 shows that property values have generally changed with changes in rents. It compares
the rate of increase in the average sales price per square foot of Berkeley properties with five
units or more built prior to 1980, the rate of increase in median rents for rent-stabilized units and
the rate of increase in median one-bedroom vacancy rents since vacancy decontrol began in
1999.31
From 1978 to 1998, just prior to vacancy decontrol, changes in Berkeley’s multifamily
property values per square foot closely tracked both changes in Berkeley’s rent levels and the
increase in the consumer price index. From 1978 to 1998 the average price per square foot went
from $31.49 to $78.44, matching the 157% increase in the CPI during the same period. In other
words, the “real” inflation-adjusted price of Berkeley rental property remained about the same
from 1978 to 1998.32
31
The initial index number for vacancy rents is based on the ratio between the median for all rents and the median
one-bedroom vacancy rent in 1999. No median is available for all vacancy registrations, but with two-thirds of all
vacancy registrations occuring in units with one bedroom or less, the one-bedroom median only slightly
underestimates the overall vacancy median. In addition, vacancies occur at a higher rate in the areas near the U.C.
Berkeley Campus that serve a predominantly student population and have somewhat higher than average rents, so a
modest underestimate is likely to better represent the rental market as a whole. 32
A study by Michael St. John covering 1970 to 1988 found much the same thing for the early period of strong
rent control. He indicates that the “real” inflation adjusted sales price per square foot of Berkeley rental property
with 5+ units had been stable during most of the 1970s, began to rise in 1978-79 and then returned to the prior “real”
values in the 1980s. Michael St. John, “The Impact of Rent Controls on Property Value”, Working Paper No. 90-
178, Center for Real Estate and Urban Economics, U. of California at Berkeley, May 1990, p.18.
26
From 1998 to 2008 the CPI increased by 36% but property values tripled, reaching $230 a square
foot and $166,500 per unit. Just under 30% of Berkeley’s pre-1980 rental properties with five
units or more have sold since 1999, meaning that 70% of these properties remain under the same
ownership and these owners have enjoyed extraordinary increases in property values.33
The rapid increases in sales prices from 1999 to 2002 reflected investor expectations that there
would be major future rent increases, since vacancy decontrol allowed rents to increase to market
rate as tenants moved, and the dot-com bubble created unusually rapid increases in market rents.
From 2003 to 2008, however, the property sales prices increases were greater than could be
explained by either actual or reasonably anticipated future rent increases. Instead, they reflected
declining interest rates and the unrealistic expectations of the housing bubble. The Federal
Reserve lowered interest rates to keep the economy from falling into a recession when the stock
market collapsed at the end of the dot-com bubble and these low interest rates helped create a
housing price bubble. Lower mortgage interest rates lowered the cost of borrowing to purchase
real estate and lower interest rates lowered the rate of return from bonds, which raised the value
of alternative investments in rental property that promised a steady or increasing net operating
income (NOI). Increases in property values resulting from the increased value to investors of
each dollar of NOI are reflected in changes in the capitalization rate, which means the NOI
divided by the sales price. Capitalization rates or “cap rates” measure the initial rate of return
that the NOI provides an investor paying a given sales price to purchase the property.
Figure 8 shows average capitalization rates for rental property sold in Berkeley from 1990 to
2009.34
33
Alameda County property data accessed through RealQuest listed 1,311 multi-family properties with five+ units
built prior to 1980, of which 384 had been sold during the period between January 1999 and December 2009. 34
CoStar Group, “City of Berkeley Apartment Sales, 5 units & greater, 1/1/1990 – 10/13/2009.
27
The financial crisis and recession that began in 2008 slowed property sales, made financing more
difficult and reduced prices. Capitalization rates rose in 2009 to 6.6%. The average price for
properties built prior to 1980 with five units or more fell to $192 per square foot and $162,500
per unit, levels that are more in line with median market rents. There were some sales in 2009
that were for less than the previous buyer had paid only a few years earlier.
Comparison of Rental Property Sales Prices in Berkeley and Alameda County
Berkeley rental property sales prices have been higher than those for Alameda County as a
whole. CoStar Group, a major commercial real estate information company, maintains a database
on property sales that goes back to 1990 and includes Berkeley and Alameda County.
Figure 9 compares the median price per square foot for sales of residential rental properties with
five units or more in Alameda County and in Berkeley, with the Berkeley prices limited to
properties built prior to 1980 in order to show any potential effects of rent stabilization. Price
data for 2009 is for the first half of the year.
During the strong rent control period, prior to the retroactive increase for maintenance of NOI in
1991, Berkeley prices were lower than those for Alameda County as a whole. After the
maintenance of NOI increase Berkeley prices were virtually identical to those for the county as a
whole until vacancy decontrol began. From 1998 on, Berkeley prices have been higher than
those for all of Alameda County.
28
Trends in the Value of Multi-family Properties by Size
Figure 10 shows the changes in average multifamily property values from 1978 to 2009 for
properties with 5 units or more, 4 units and 2 units.35
Two-unit properties are partly in the homeownership market and partly in the rental market, with
the result that the average price per square foot is substantially higher than in properties with 5 or
more units, which are almost exclusively rental properties. Two-unit properties are also much
more responsive to increases in prices in the homeownership market, as is shown in the increase
in the price of two-unit properties in the late 1980s and during the housing bubble that began in
the Bay Area with the “dot-com” boom in 1998 and declined after 2006.
The prices for four-unit properties were generally similar to those for properties with five or
more units from 1978 to 1986 and from 1991 to 1997. In 1986 - 1991 and 1997 – 2003 they were
somewhat higher priced than properties with five units or more. The late 1980s was a period of
rising home prices in which there was extensive conversion of small rental properties to
35
Alameda County property records, through RealQuest.
29
homeownership through tenancy in common (TICs).36
During the five year period from 2004 –
2008, the height of the housing bubble, prices rose rapidly to levels that can only be explained by
people investing in four-unit properties to use them for owner-occupancy or sell them to owner-
occupants. (Conversion to owner-occupancy can be done through conversion to condominiums,
sale of tenancies in common or converting a property that was originally a single-family home
back from four units to the original one.) The financial crisis took full effect in 2009. As home
prices declined and unconventional financing became more difficult to find, four-unit property
values returned to the level of other rental properties.
The prices of properties with five units or more increased slowly until 1998 when both vacancy
decontrol and the “dot-com” bubble began. After 1999 prices rose rapidly and then, in 2009, fell
back to the level of prices in 2003. Differences between 5-9 unit properties and those with 10
units or more were minor so we simply show all those with five units or more together.
The Effect of Vacancy Decontrol on Property Values
Given the general tendency for trends in multifamily property values to follow trends in rents, we
can estimate the effect of vacancy decontrol on property values. We compare current values with
the expected values if Berkeley had a system of strong rent controls that indexed rents to the rate
of inflation. If prices since 1998 had continued to follow the increase in the CPI, as they did from
1978 to 1998, then as shown in Figure 6, by 2009 they would have reached only $83,000 per unit
rather than $162,500 and $111 a square foot rather than 2009’s $192.
We can produce a very conservative estimate of the total value of all 19,000 rent stabilized units
by simply applying these average prices and ignoring the greater value of many 2 – 4 unit
properties. On a per unit basis the total value of all of these units is approximately $3.1 billion,
with $1.5 billion added by vacancy decontrol.
Making an estimate based on average per square foot prices is more complex. The RealQuest
compilation of Alameda County data indicates that there are nearly 11 million square feet of
multifamily residential property in Berkeley in 15,425 units in properties with 5 or more units for
an average size of just over 700 square feet. This would yield an average per unit value of
$135,000 under vacancy decontrol and $78,000 under strong rent control. Applied to all 19,000
units this results in an estimate of current total value of $2.6 billion, increased by $1.1 billion by
vacancy decontrol. The discrepancy between per unit and per square foot values may indicate
that the properties sold in the base year 1978 tended to be smaller than average or that those sold
in 2009 tended to be somewhat larger or it may simply reflect the limits of the data that is kept
by the Alameda County for tax purposes.
We have estimated a total annual rent roll of $285 million in registered units. Applying an
average NOI of 60% and a capitalization rate of 6.6% results in an estimated total current value
of $2.6 billion. The estimated rent roll for a hypothetical strong rent control system was $184
million. Applying an average NOI of 50% and a capitalization rate of 6.6% results in an
36
Michael St. John, “The Impact of Rent Controls on Property Value”, Working Paper No. 90-178, Center for
Real Estate and Urban Economics, University of California at Berkeley, May 1990, pp.25-29.
30
estimated total value of $1.4 billion. This suggests that vacancy decontrol has increased property
values by $1.2 billion, within the range of $1.1 to $1.5 billion in the previous estimates.
The Effect of Vacancy Decontrol on City Tax Revenue
The increased rents and sales prices of property generate increased tax revenue for the City of
Berkeley through the property tax, real property transfer tax and the business license tax and
decreased revenue through the sales tax.
Real Property Transfer Tax
Over the past eleven years since vacancy decontrol nearly 30% of Berkeley’s rental properties
were sold, a rate just under 3% annually. Based on the 1.5% transfer tax rate and the estimate
that rental property values have increased by from $1.2 billion, the increased annual tax revenue
would currently average around $450,000.
Business License Tax on Gross Rent
The tax rate is 1.08% so based on the estimated annual rent increase of $100 million, annual tax
revenue should have increased by $1,080,000. (Assumes full compliance. The Auditor has been
finding significant underpayment.)
Property Tax
The basic property tax rate is 1%. If all properties were reassessed at new, higher values, that
would increase overall County property tax by $12 million a year. However, so far about 30% of
multi-family rental properties with five units or more have been sold since vacancy decontrol,
and many of those were sold at prices below current market values. We estimate that so far
County tax revenues have increased by $2,250,000, of which 32.5% or $730,000 goes to the
City. The remainder goes to the Berkeley Unified School District, the Peralta Community
College District, Alameda County and other special purpose districts.
Sales Tax
Berkeley receives 1% of the sales tax collected here. All tenants live in Berkeley and many do
not have cars and spend most of their money in Berkeley, so the maximum possible decrease in
sales tax revenue would be $1 million. Not all money going to increased rent would be spent on
taxable items in Berkeley however. In addition, while the majority of investors in rental property
live outside of Berkeley a substantial minority live in Berkeley. A more plausible guestimate
would be a $300,000 reduction in sales tax revenue. In addition, there is an unknown loss of
business license tax from retail businesses due to the reduced sales.
Total
Overall, vacancy decontrol has likely increased tax payments to local governments in Alameda
County by about $3.5 million annually, including an increase in City of Berkeley revenues of
nearly $2 million annually. This does not take into account costs that the City or County may
incur due to increased financial hardship to tenants.
31
Price, Leverage and Risk
The value of rental property, whether in Berkeley or anywhere else in the U.S., is based on its net
operating income (NOI), which is the flow of income that an investment in that property
provides the investor after paying the operating expenses. The average NOI reported for
properties sold during the past 10 years in Berkeley is 62% and for Alameda County is 61%.37
Typically there are two investors in a rental property, an owner and a mortgage lender. The
mortgage holder receives payments from the NOI first and the loan is secured by the right to
foreclose on the mortgage and take over ownership if these payments are not made. In return for
this greater security, the mortgage lender accepts a limited return with either a fixed or variable
interest rate. Owners can profit from an investment in rental property in four ways: The owners
gain equity as the value of the property increases, they gain equity as any mortgage used to help
purchase the property is amortized; the owners gain positive cash flow from the NOI (less
investment costs such as mortgage repayment) and the depreciation allowance provides the
owners with tax savings that increase their after-tax cash flow. All of these depend on NOI.
Investors may use shortcuts such as a gross rent multiplier, but prudent investors will use such
shortcuts only to screen properties for further investigation.
Investors have to choose how much of the purchase price they want to pay with their own money
and how much they want to borrow. If the investor purchases the property using only his or her
own money, then the investor receives the entirety of the NOI as positive cash flow as well as
getting an equity gain to the extent that the property increases in value at the time of sale. The
more the investor borrows, the greater the “leverage,” which means that the potential rate of
return on investment is higher because the investor uses less of his or her own money and still
gets all the equity gain from increased property values. In addition, capital gains from increased
property values are generally taxed at a lower rate than corporate or personal income from
positive cash flow.
Highly leveraged investments have higher risk for the investor. First, they put most of the NOI
into repaying a mortgage. This reduces the investor’s cash flow and creates a risk of foreclosure
if the investor is unable to keep up the payments due to an increase in expenses or a decline in
rents. Second, more of the investor’s return on investment depends on increased property values,
which depend not only on increased NOI but also on capitalization rates, which usually track
interest rates and can change unpredictably from year to year. The present value of any
particular property will vary depending on the current and expected future NOI and interest rates.
The value of the property increases with the perceived likelihood that NOI will increase in the
future. The value of the property also increases if the current and expected rates of return for
alternative safe investments such as government bonds are expected to decrease or remain low.
37
CoStar, “City of Berkeley Apartment Sales, 5 Units & Greater, 1/1/1990 – 10/13/2009, built 1979 & older”.
CoStar, “Alameda County, Apartment Sales, 5 Units & Greater, 1/1/1990 – 10/13/2009”.
IREM Institute of Real Estate Management, "Income/Expense Analysis: Conventional Apartments", 2008
reports a median NOI of 62% in 2007for garden apartment buildings in the Oakland metropolitan area.
32
As the property turns over, each buyer pays the seller the capitalized value of the NOI and
possibly also of some of the future increases that the buyer expects will take place. Normal
mortgage financing will use up to 80% of the NOI for mortgage payments and reduce the new
owner’s cash flow to a modest level. Similarly, as longer-term investors refinance to take out
accumulated equity they also take on mortgage obligations that reduce cash flow to a modest
level. Many things may not go as the investor anticipated: increases in rents may not be as fast
as expected; operating and maintenance costs may be higher than expected; interest rates may
increase on a variable rate mortgage, among other possibilities. These further reduced cash flow
and create an incentive for the owner to reduce maintenance in an effort to recapture their
investment.
The market for investments is generally biased in favor of increased risk. When an investor in
rental housing decides to sell, the property is sold to the highest bidder. That is another way of
saying that the property is sold to the bidder who makes the most optimistic assumptions about
future rents, operating, maintenance and renovation costs, net operating income and interest
rates. It is not uncommon for buildings to be purchased for prices that are based on mistaken
expectations about future rent increases rather than on current rents. The Federal and State tax
systems reinforce this bias towards risk by taxing capital gains at lower rates than income. To the
extent that investment in rental property is highly leveraged and based on expectations of
substantial future increases in net operating income the investment becomes more speculative
and carries serious risks for the community as well as for the investor.
There is excellent evidence for this phenomenon in a survey of landlords in Los Angeles,
conducted in November 2007 through February 2008. During the 2000s, rents in Los Angeles
rose at an exceptional pace by national standards. The median contract rent went from $612 in
2000 to $933 in 2008.38
Yet only 28% of apartment owners reported making a profit in the
previous year. Fully one-third of respondents reported a loss in the previous year, another quarter
reported that they only broke even and another 14% did not know how they did.39
The majority
of Los Angeles landlords who purchased their properties prior to 2000, before the housing
bubble, reported that their property was profitable, while the vast majority of those who
purchased after 1999 reported that their property was not profitable.40
In the L.A. study 23% of
landlords reported that they were postponing maintenance on major problems rather than
handling them as quickly as possible, and an equal proportion reported that they were postponing
maintenance on minor problems.41
So despite rent increases of over 50% in only a few years,
many landlords in Los Angeles are running negative cash flows due to the size of the mortgages
they have taken on in order to purchase the properties or in order to take out equity from their
increased property values and now they are failing to properly maintain their properties.
How much the current owner profits from Berkeley’s increasing rents and property values
depends on when they bought the property, how much they paid for it, how they financed the
purchase, how long they hold it and whether rents and property values increase at the expected
38
U.S. Census Bureau. 2000 Census and 2006-8 American Community Survey. 39
Economic Roundtable, Economic Study of the Rent Stabilization Ordinance (RSO) and the Los Angeles Housing
Market, Los Angeles, 2009, pp. 195-196. 40
Ibid, p.199. 41
Ibid, p.192.
33
rate. Approximately 30% of Berkeley rental properties with five units or more were purchased
since 1999, during the series of housing-related economic bubbles, and an unknown number
have refinanced.
The revenue side of most investment in Berkeley’s rental property is not very risky as long as
owners do not overestimate the rate of future increases. The presence of the University of
California at Berkeley guarantees high demand for housing within walking and bicycling
distance from Campus, an area encompassing the majority of Berkeley’s rental units, and the
University’s investment in seismic reinforcement of its buildings ensures that even in the event
of a major earthquake the University will be back in operation in a fairly short period of time.
Competition to meet this demand is limited. The city’s relatively high densities mean that
increases in supply in supply within the city can only be made in an already built-up area where
the cost of constructing additional housing will be high. Overly high expectations for future
increases, however, may have negative effects on building maintenance, especially if rents level
off or even decline for a few years as a result of the current recession.
There are also risks in projecting future costs, especially with buildings that require seismic
reinforcement or where major building systems are coming to the end of their useful life and
need major renovations or where other requirements, such as energy efficiency, are strengthened.
Recommendations:
The City should investigate measures to reduce the adverse impacts of the speculative element of
investment in rental property in Berkeley. Such measures might include requirements that take
effect when a rental property is transferred or refinanced:
Mandatory code inspection prior to or no more than 30 days after the time of sale or
refinance and all code violations cleared.
Energy efficiency inspection prior to or within 30 days of the time of sale or refinance and
must meet required standards (possibly as set by RECO, CECO).
Soft story buildings fully retrofit to meet a life-safety standard within one year of transfer.
34
Section 4: Fair Return and Low-Turnover Properties
Fair Return in Berkeley
Rent control systems must provide for increases in the legal rent ceiling that allow sufficient
increases in Net Operating Income (NOI) so that the owner receives a “fair return” on investment
and the return is not gradually reduced by inflation. (The analysis of a fair return applies to a
property as a whole, not to individual units within a property, and there can be considerable
variation in allowable rents among different units as long as the property as a whole provides a
fair return.) NOI increases may be set at less than a full inflation adjustment, because a normal
amount of leveraging will still allow an owner to maintain their current level of profitability. The
following example illustrates this point.
Table 9: Example of the Effects of Leverage on Cash Flow Under Rent Control
Base Year Current Year Increase over Base Year
Consumer Price Index 100 125 25%
Gross Rental Income $10,000 $12,000 20%
Operating Expenses -$4,000 -$5,000 25%
Net Operating Income $6,000 $7,000 17%
Debt Service Payments -2,000 -2,000 0%
Cash Flow $4,000 $5,000 25%
Table 9 simulates the effect of a rent control annual increase formula that set the allowable
increase at 80% of the increase in the Consumer Price Index in a property with no turnover
among its tenants. With a CPI increase of 25%, rents increase by 20%. With operating expenses
assumed to increase by the rate of inflation, 25%, the NOI increases by only 17%. Although this
is an increase in total dollars, it is a decline in real inflation-adjusted dollars to the equivalent of
$5,600 in base year dollars. As a proportion of rental income NOI declines from 60% to 58%.
Yet the owners’ cash flow increases by 25%, the same as the increase in the CPI, because the
debt service payments were fixed. This example uses a very conservative level of debt service,
since lenders will typically allow debt service to reach 70% of the NOI. If, for example, the debt
service payments were $4,000 and the initial cash flow $2,000, then the increase in cash flow
would be 50%, double the rate of the CPI. This does not take into account the additional profit
resulting from amortization of the mortgage and the tax savings from the depreciation allowance.
Table 10 shows the same example, but using an AGA set at 65% of the CPI. In this hypothetical
case, the increase in cash flow does not keep up with inflation, although it could with a higher
degree of leveraging or if some tenants moved during the year so that vacancy increases raised
total rental income.
35
Table 10: Example of the Effects of Leverage on Cash Flow with 65% of CPI AGA
Base Year Current Year Increase over Base Year
Consumer Price Index 100 125 25%
Gross Rental Income $10,000 $11,625 16%
Operating Expenses $4,000 $5,000 25%
Net Operating Income $6,000 $6,625 10%
Debt Service Payments -2,000 -2,000 0 %
Cash Flow $4,000 $4,625 16%
We have previously seen that the average NOI in Berkeley is likely around 60%, which means
that operating expenses constitute the other 40% of the rent. The Rent Board has established an
NOI adjustment of 65% of CPI as sufficient to ensure that most owners are able to maintain a
fair rate of return on their investment.42
If we apply this standard to the 60% of the rent that goes
to NOI, then a rent increase of 39% of the increase in the CPI is necessary to provide the NOI
adjustment (.65X.6=.39). Operating expenses are the other 40% of the rent, and there is a
“historical tendency for apartment operating costs to track the general rate of inflation.”43
If
operating expenses increase at the same rate as the CPI, then a rent increase of 40% of the
increase in the CPI is necessary to meet that increase. In total, then, the annual rent increase
called for in this example will be 79% of the CPI.
The best available historical evidence on actual changes in operating costs is from a combination
of the annual cost studies carried out for the Berkeley Rent Board, which provide limited
information but are specifically about Berkeley, and the annual apartment operating cost surveys
published by the Institute of Real Estate Management, which provides more comprehensive data
on operating costs for the broader Oakland metropolitan area.
In Berkeley annual operating cost studies were done from 1980 to 2004 for purposes of setting
the AGAs from 1981 to 2005. The 1980 and 1981 studies were very preliminary, but we have a
series of comparable studies covering changes in operating costs for the 23-year period from
1981 to 2004. These studies provided the basis for AGAs from 1982 to 2005. These studies
assume a property in which the annual rent increases for all units have been limited to the AGA,
ignoring any vacancy increases that have occurred since 1999. The studies are based on the
actual changes in the cost of taxes, fees and utilities, but not on actual changes in the costs of
management, maintenance or insurance. Instead they assume that these later three expense
categories increase with inflation.
The 1982 study was re-examined and modified slightly in 1991, and we will use the modified
version as the baseline for comparison with the 2004 study to look at changes in operating
42
Based on Kenneth K. Baar, “Issues and Options for Rent Increase Standards Under Berkeley’s Rent
Stabilization Ordinance”, October 26, 1995. 43
Hamilton, Rabinovitz & Alschuler, The 1994 Los Angeles Rental Housing Study: Technical Report on Issues
and Policy Options, December 1994, p.245. Prepared for the Los Angeles Rent Stabilization Board.
36
expenses over time.44
(This analysis looks only at individually metered units in which the tenant
pays for gas and electricity used within the unit and the owner pays gas and electricity only for
common areas, the predominant format in Berkeley.) During this period the Bay Area CPI
increased by 119%, and according to the studies operating expenses increased by 197% (a “real”
inflation-adjusted increase of 36%) By far the largest source of these increases was local
government efforts to deal with the loss of property tax revenue due to Proposition 13 by
establishing local parcel taxes.
Table 11: Changes in Berkeley Operating Expenses, 1981 to 2004
1981 2004
Median Rent $234 $602
Operating Expenses $103 $306
OE as percent of Rent 44.1% 50.8%
The Rent Board’s annual cost studies simply assumed that the costs of management,
maintenance and insurance would increase at the same rate as the CPI and there is no source of
data on these costs specifically for Berkeley. However the Institute of Real Estate Management
of the National Association of Realtors (IREM) publishes an annual report for the United States
and major metropolitan areas based on an annual membership survey, and this provides data on
the Oakland area as well as on the U.S. as a whole. The IREM data comes primarily from large
apartment complexes with an average of over 100 units, so its specific cost findings are not
directly applicable to most of the Berkeley market, but it does provide information on trends in
operating costs in the East Bay. IREM has used a consistent methodology from 1985 to the
present.45
Table 12: Changes in Oakland Area Cost of Administration, Maintenance & Insurance as
Reported by IREM Members
1985 2006 Increase
Admin., Maint. & Ins. $1.79 $2.81 57%
Bay Area CPI Index 108.4 209.2 93% Source: U.S. Bureau of Labor Statistics; Institute of Real Estate Management of the National Association of
Realtors, Income/Expense Analysis: Conventional Apartments, IREM, Chicago, 1986 and 2007 editions.
44
Hamilton, Rabinovitz & Alschuler, “Inflation Indexing in Berkeley Rent Regulation in the Aftermath of the
Searle Decision”, August 19, 1991, pp.22-26. Prepared for the Berkeley Rent Stabilization Board.
Kenneth K. Baar, “Apartment Operating Cost Increases in Berkeley: Analysis for the 2005 Annual General
Adjustment” presented to the Berkeley Rent Stabilization Board, October 18, 2004. 45
Institute of Real Estate Management of the National Association of Realtors, Income/Expense Analysis:
Conventional Apartments, IREM, Chicago, 1986 and 2007.
37
As shown in Table 12, the IREM surveys indicate that from 1985 to 2006 the cost of
administration, maintenance and insurance in the Oakland area increased at only 81% of the rate
of increase in the CPI. Nationally they increased by 91% of the CPI. This suggests that the
presumption that these costs increase at the same rate as the CPI is somewhat conservative and
may result in overestimating the actual increase in operating expenses.
The 2004 Berkeley cost study estimates that in the average property that has never received a
vacancy increase the operating expenses and NOI each make up half of the rent. If the NOI
maintenance standard is that it must increase by 65% of the rate of increase in the CPI, then an
increase in the rent of 32.5% of the CPI increase is necessary for maintenance of NOI, since NOI
is half of the rent. If operating expenses increase by 135% of the rate of the CPI, then a rent
increase of 67.5% of the increase in the CPI is necessary to cover increased operating costs. The
two add up to rent increase of 100% of the CPI.
Berkeley’s annual cost studies from 1981 to 2004 found that rent increases roughly equal to the
increase in the CPI-All Items have been necessary to compensate for increases in operating costs
and provide an adequate increase in NOI in order to provide a fair return in buildings without
tenant turnover or under strong rent control where units do not decontrol on vacancy.46
Past
history does not mean the future will follow the same pattern, but over all we can say that the
AGAs from 1980 to 2005 clearly met constitutional fair return standards and that they provided
rent increases roughly equal to the increase in the CPI-All Items during that period.
Effects of Tenant Turnover Rates on Rents
Berkeley’s rent stabilization system is required to ensure that owners receive a fair return on
their investment. Properties subject to rent stabilization will be affected differently based on a
combination of the rate of turnover among their tenants and the number of exempt units.
Properties where all units either have turned over since 1998 or are rented to tenants in the
Section 8 program will have rents that are at or close to current market levels and have received
rent increases well above the U.S. average. Properties where all tenants have been in place
continuously since prior to January 1, 1999 have received the Annual General Adjustments,
which until 2005 were based on annual cost studies to ensure a fair return. Many properties have
a mixture of both long-term and post-vacancy decontrol tenancies.
As of September 2009, 78% of registered rental units had reported a new tenancy since January
1, 1999, nearly 11 years ago, while 22% had not. Vacancy increases are not evenly spread out
among all Berkeley rental properties. They are more frequent in the areas close to the UC
Berkeley campus, where the student population is most concentrated. Larger properties generally
have substantial turnover, partly because more of them are located near campus and partly
because with more units each property is more likely to have an experience that reflects average
turnover rates.
46
By 2005 the Rent Board’s consultant reported that “excess” NOI increase that resulted from adjusting NOI by
100% of CPI in 1991 rather than 65% of CPI had been almost entirely eliminated by reductions in subsequent
annual increases. Memorandum dated October 18, 2004 from Jay Kelekian, Executive Director, to Members of the
Rent Stabilization Board, “2005 Annual General Adjustment”, pp.4-5.
38
According to the currently available information, there are 271 units in 186 properties where no
unit has received a vacancy increase. These are mostly in small properties, including 98 single-
family or condominium units, another 98 units in 49 two-unit properties and 75 units in
properties with three units or more.
There are another 472 units that have not received a vacancy increase in 59 properties where at
least one other unit has turned over, but where at least two-thirds of the units have not received a
vacancy increase.
In total, then, there are 643 units in 245 properties, less than 4% of all registered rental units,
where two-thirds or more of the units are not reported to have received any vacancy increase.
The current AGA formula, in effect since 2005, sets annual rent increases for sitting tenants at
65% of the increase in the Bay Area Consumer Price Index for All Urban Consumers during the
June to June period of the previous year. During the period from June 2005 to June 2009 the Bay
Area CPI-All Items increased by 12.3%, so that under the formula, AGAs as of January 2010
have increased rent ceilings by 8.0%. This means that in those properties where all or nearly all
tenancies began prior to 1999 the increases have been 4.3% less than the increase in the CPI.
This will not be a problem in buildings where there has been turnover. The average rent in
Berkeley's rent stabilized units overall increased by 18% between 2005 and 2008, well above the
rate of inflation.
In buildings where there is little or no turnover the disparity will increase over time and may be
problematic for some owners. This could lead to petitions for individual rent adjustments based
on net operating income that would require an administratively burdensome level of analysis for
the Rent Board as well as additional costs for the owner. The administrative costs to both the
Rent Board and the owner could be reduced by creating an expedited Individual Rent
Adjustment process that simply assumes the historical rate of cost increases and allows increases
in the rent ceilings sufficient to increase the property’s total rent by the increase in the CPI since
2005.
Recommendation
Provide an expedited Individual Rent Adjustment process that allows increases in the rent
ceiling for properties where the aggregate AGAs for all units in the property have not provided a
rent adjustment that matches the increase in the Bay Area CPI-All Items.
39
Section 5: Changes in Rent Burden, 1980 – 2009
Bay Area Rent Burden, 1980 - 2008
The Bay Area’s extraordinary rent increases have had major consequences for the quality of life
available to most low-income. Fully 25% of Bay Area tenants, and virtually all very low-income
tenants who do not receive rent subsidies or live in non-profit housing, pay half or more of their
income for rent.47
Figure 11 shows the incomes of tenants at the 25th
percentile, the lowest-
income quarter of all tenants, and shows how much of that income would be required to rent a
unit at the 25th
percentile, the lowest quarter of all rents.48
The figure first compares incomes and
rents for low-income tenants in 2008 with 1980, just after Berkeley, San Francisco, San Jose and
Oakland instituted rent stabilization. It then shows what the Bay Area tenant income after rent
would be if rent increases had been held to the increase in the Consumer Price Index so that the
“real” rent did not increase.
47
2006-2008 American Community Survey 3-Year Estimates, Gross Rent as a Percentage of Household Income. 48
Sources: 1980 Census, 2006-2008 American Community Survey, Bureau of Labor Statistics, author’s
calculations. (1980 census data provides rent for 1980 and income for 1979. Both are adjusted to 2008 using the
CPI-Less Shelter, since the purpose of the exercise is to measure buying power left after paying rent.)
40
Not only have real incomes gone down for these tenants, real rents have increased so much that
in 2008 this group of tenants has one-third less money left after paying rent than they had in
1980. The percentage of income that a tenant at the 25th
percentile in income would have to pay
for a unit whose rent was at the 25th
percentile went up from 28% in 1980 to 48% in 2008.
Figure 12 shows that incomes have gone up since 1980 for median-income tenants, but increased
rents have eaten up the entirety of the increase.49
The amount of income required for the median
tenant to rent the median unit has increased from 20% in 1980 to 30% in 2008.
49
Source: 1980 Census, 2006-2008 American Community Survey, Bureau of Labor Statistics, author’s
calculations. (1980 census data provides rent for 1980 and income for 1979. Both are adjusted to 2008 using the
CPI-Less Shelter, since the purpose of the exercise is to measure buying power left after paying rent.)
41
Berkeley Rent Burden, 1988 - 2009
Determining the effect of rising rent on tenants in Berkeley requires taking into account the
city’s large student population. The Rent Board commissioned surveys of tenants in rent
stabilized units in 1988, 1998 and 2009 in which non-student households’ incomes and rent
burdens are analyzed separately, so we can use this to track the rent burdens of the non-student
households over time. In 1988 the median rent burden for non-student households was 22%.
This increased to 25% in 1998 and it remains at that level for pre-vacancy decontrol tenants, but
for more recent non-student tenants the median rent burden is now 32%.
Table 13: Rent Burden of Non-Student Households: 1988, 1998, 2009
Survey Year 1988 1998 2009
All Pre-1999 Post-1998
Up to 30% income 71% 59% 50% 58% 47%
More than 30% 29% 41% 50% 42% 53%
More than 50% 14% 20% 26% 21% 28%
In 2009 26% of all non-student tenant households reported that they are severely rent burdened,
paying over 50% of their income for rent and utilities. In comparison, in 1998 20% paid over
50% of income for rent and only 14% in 1988. This means approximately 3,400 non-student
households were severely rent burdened.
Looking separately at pre- and post-vacancy decontrol tenants we find that among post-vacancy
decontrol non-student households the median rent burden is 32%, with 53% of households
paying over the Federal affordability standard of 30% of income for rent and utilities. Twenty-
eight percent paid over 50% of their income in gross rent. This projects out to a total of 5,000
rent burdened households in this group, including about 2,650 households that are severely rent
burdened.
Overall, the situation in Berkeley is not very different from that in the Bay Area as a whole, but
it is significantly better for long-term tenants. For pre-vacancy decontrol tenancies the median
rent burden is 25%, with 58% paying no more than 30% of income for rent and utilities and 42%
paying over 30%. One fifth (21%) pay over 50% of income in gross rent. There are
approximately 3,500 pre-vacancy decontrol tenancies in Berkeley, so this indicates that
approximately 1,500 households are overpaying for rent despite the benefits of continued strong
rent controls, and about 750 are severely overpaying. The rent burden of these long-term non-
student tenants remains basically the same as it was for all tenants in 1998.
If rent controls were eliminated the rents of long-term tenants would be increased by at least
50%. Such a rent increase would raise the median rent burden for this group to 36% of income.
The proportion of households paying over 30% of income for rent would increase to 59%, with
more than half of those (36%) paying over 50% of income for rent.
42
Section 6: The Economics of Rent Stabilization
Berkeley’s Rent Stabilization and Eviction for Good Cause Ordinance states that “The purposes
of this chapter are to regulate residential rent increases in the city of Berkeley and to protect
tenants from unwarranted rent increases and arbitrary, discriminatory, or retaliatory evictions, in
order to help maintain the diversity of the Berkeley community and to ensure compliance with
legal obligations relating to the rental of housing. This legislation is designed to address the City
of Berkeley's housing crisis, preserve the public peace, health and safety, and advance the
housing policies of the city with regard to low and fixed income persons, minorities, students,
handicapped, and the aged. “
In order to properly implement the ordinance, the Rent Board has defined when rent increases
are “unwarranted” through its implementing regulations. These regulations that start with 1980
base rents and then allow rent increases to cover increases in operating costs and, where
renovations require a significant capital investment, increases to cover both the costs of the
renovations and a reasonable interest rate on the investment, as well as increases for net
operating income set at 65% of the increase in the consumer price index. As we have seen, this
formula would generally have allowed rents to increase at roughly the same rate as the consumer
price index over the past thirty years since rent stabilization began in Berkeley.
Opponents of rent control argue that rent controls are inherently destructive because any
reduction in market rent will result in a corresponding decrease in the quality and quantity of
rental housing. Yet we saw in Table 4 that rents are substantially lower in most U.S.
metropolitan areas than they are in the Bay Area and these areas continue to produce and
maintain decent quality rental housing. Portland is a thriving city and metropolitan area where
the median rent is 61% of the median rent in the San Francisco Bay Area. How is it that
landlords in Portland are able to profitably operate and maintain rental housing whose quality
seems quite comparable to rental housing in the Bay Area while charging rents that are
comparable to what Berkeley rents would have been under strong rent control?
The critics of rent control assume an ideal market. In a completely open and competitive market,
competition among the producers of goods and services holds housing costs to the minimum
feasible level and provides its own stringent price controls. Under ideal conditions, if demand
increases, then the price will increase, but only for a short period of time. Higher prices result in
higher profits that bring in additional investment in production of housing and this increases the
supply. The increased supply increases competition among sellers and brings the price back
down. In this ideal housing market all prices are either at the minimum necessary to provide a
desired good or service or will be reduced to that minimum soon. If prices in this fully
competitive market are reduced by regulation, then the reduction in profitability will lead to a
reduced supply as producers shift their investments to other more profitable opportunities and the
reduction in cost will result in inefficient allocation of units as consumers lucky enough to obtain
housing at below market rate stay in apartments that they would otherwise leave and make
available to other tenants. This is the standard economic critique of rent control.
In the real world, however, many markets have substantial and durable barriers to competition
that allow businesses to raise prices and extract unearned “windfall” profits from consumers over
43
a long period of time. This is why there is public regulation of the prices charged by private
companies that distribute electricity, gas and water, for example, and it is one of the major
reasons for rent regulation. Rental housing is not a “natural” monopoly, the way distribution of
gas or electricity is, but even though there are many separate owners of residential rental
property there are significant barriers to adequate competition in many rental housing markets.
Production of new rental housing is a separate business from operation and maintenance of
existing rental housing and rental housing production requires land that is suitably located and
zoned for construction of multi-family housing.
In a fully competitive housing market production of housing can easily be increased because
sites to build on are readily available and the cost of the land is minimal. In many successful
urban areas, however, there are physical constraints resulting from the density of the built
environment that make it more difficult to find suitable sites and more costly to build on those
sites. Land use regulations also add to the difficulty of finding sites on which to build more
housing, especially multi-family housing in the already developed areas. And much of the land
near the central Bay Area is occupied by water or steep slopes that make it very difficult to build
housing.
In addition, housing is a product that requires a major up-front investment, another similarity to
utilities. Once the costs of construction are paid, usually over a significant period of time (the
standard amortization period for tax purposes is 27.5 years) the cost of operating and maintaining
the housing is greatly reduced. If an urban area has long-term stability in population and
household formation this will likely result in an adequate supply of older housing that will be
affordable to many renters with below-average incomes. However, where population and
household formation increase faster than increases in multi-family housing production, then the
scarcity of older housing drives up rents to levels closer to the cost of new rental housing. As a
result of these two factors, scarcity of sites for development and mismatch between past rates of
housing production and current need, housing in many economically successful urban areas has
prices far greater than would be charged in a fully competitive market.50
This revenue over and above the economically necessary costs of producing, operating and
maintaining the housing plus a reasonable return on the investment is what economists call
“economic rent”, meaning revenue that is over and above what is necessary to support
production of a good or service. Economic rent is not earned through production but rather is
unearned revenue from ownership of a needed commodity, such as places to live, in a setting
where demand is higher than supply. This is typical of successful urban areas in coastal settings
where the limited supply of additional developable land limits competition. Economic rent in the
housing market is typically called land rent, because payments over and above those necessary to
operate and maintain the building are attributed to the land or location.
The San Francisco Bay Area has high demand driven by a successful knowledge-based
economy, major investments in education, transportation and other public services and the high
50
For an overview of various types of market failures in the rental housing market ,including the historical
mismatch issue, see Andrejz Skaburskis and Michael B. Teitz, “The Economics of Rent Regulation”, pp.41-60 in W.
Dennis Keating, Michael B. Teitz and Andrejs Skaburskis (editors), Rent Control: Regulation and the Rental
Housing Market, Center for Urban Policy Research, Rutgers, New Brunswick, N.J., 1998.
44
quality of its regional environment and culture. It has extreme constraints on increased housing
supply in its central core and due to the combination of geographical constraints from the ocean,
the Bay and the hills, density constraints in its central core due to the high costs of in-fill
development at higher densities, and regulatory constraints on development of multi-family
housing, particularly in the outlying suburbs. For the past fifty years the amount of new rental
housing built has fallen far short of the quantity necessary to hold down rents. As a result, Bay
Area rents are far above the level necessary in order to operate and maintain rental property,
which is what rents would be in a genuinely competitive market. The Bay Area median rent is
75% higher than that for all U.S. metropolitan areas and 65% higher than the Portland, Oregon
area. About one-third of the aggregate rent paid in the Bay Area is over and above the necessary
cost of operating and maintaining the buildings. (For a detailed analysis see Appendix 4:
Estimating Land Rent in the Bay Area Rental Housing Market). Rents here are limited less by
competition and more by the limits of what tenants can manage to pay.
The presence of land rent or economic rent has major implications for the economics of different
housing policies. Prof. Lee Friedman, at the University of California’s Goldman School of Public
Policy, has demonstrated that in the presence of substantial land rent, the economic models of the
effects of rent controls become indeterminate and “perfect rent control could, in theory, affect
only economic rents and cause no supply inefficiency even in the long run”. 51
The Berkeley
Rent Board’s definition of an “unwarranted” rent increase is basically an increase in “economic
rent” or “land rent”, while allowing increases in rent that are necessary for the maintenance and
operation of the building itself plus a reasonable profit on operating the building. Mayer and
Olsen have argued that well-designed rent controls could improve maintenance and the Berkeley
Rent Board’s regulations allowing increases for maintenance costs and decreases for reductions
in service due to inadequate maintenance follow the pattern that they recommend.52
If we measure a “fair rent” and a “fair return on investment” by the rents and rate of return that a
housing market would provide under conditions of perfect competition, the evidence shows that
prior to the imposition of vacancy decontrol by the State of California Berkeley’s rent control
system had evolved, after lengthy court battles and often bitter political conflicts, into a system
that for the most part provided fair rents to tenants and a fair return on investment to owners.
Investors in rental property would have received a fair return on their investment in Berkeley
even if strong rent control had continued to the present day because the controlled rent level
would provide investors with the same rental income they would receive in a genuinely
competitive rental housing market.
Today, Berkeley’s rent stabilization program provides most tenants with valuable stability and
somewhat lower rents if they stay in place for more than a year or two. Together with “eviction
for good cause” it reduces the displacement caused by rent increases that are not justified by
51
Lee S. Friedman, The Microeconomics of Public Policy Analysis, Princeton University Press, 2002, p.537.
Friedman provides an excellent overview of the economics of rent controls in areas with substantial land rents in
Chapter 13, “The Control of Prices to Achieve Equity in Specific Markets”. 52
Mayer, Neil S. (1984) “Conserving Rental Housing: A Policy Analysis”, Journal of the American Planning
Association, 50:311-325.
Olsen, Edgar O. (1988). “What Do Economists Know About the Effect of Rent Control on Housing Maintenance”,
Journal of Real Estate Finance and Economics, 1:295-307.
45
actual increases in costs or quality of services. The ordinance successfully meets its goal of
reducing arbitrary evictions and it prevents unwarranted rent increases for tenants after they have
moved into a unit. The state requirement of decontrol on vacancy means that the Rent Board
cannot prevent unwarranted rent increases on units at the time tenants move in. As a result, only
a diminishing number of very long-term tenants are shielded from overall rents that are well
above the levels that would otherwise be considered “warranted” by the ordinance and
implementing regulations.
Residential land in Berkeley is valuable in part due to major public investments such as the
University of California, the East Bay Regional Park District, the freeways and BART and in
part due to the efforts of all those who live and work in Berkeley and give the city its character as
a center for learning and creativity. Berkeley is the home or workplace of many people who
contribute greatly to the community even though they do not make much money and do not own
their own homes. They work as attendants to people with disabilities, they work for small non-
profit service and advocacy organizations, they work in restaurants and stores, they do research
and write books that don’t make the best seller list and they do many other things that make
Berkeley a great place to be. Together the public at large and the Berkeley community make the
city a desirable place to live and work, and the result of making it desirable is that land values in
Berkeley increase and rents and home prices go up. For those who already own real estate this
often works out well, but for those who don’t it creates a cruel irony. Their own contributions to
this community make it harder for them to afford to live here.
Berkeley’s Rent Board is on record supporting a change in state law to give back to local
governments the power to regulate rents on a continuous basis could allow Berkeley restart
strong rent control again at current rent levels and prevent drastic future increases in rents. There
are many intermediate possibilities between the full vacancy control system Berkeley had in
1995 and the current modest vacancy decontrol-recontrol system that State law allows today.
Limited vacancy increases such as those allowed during the three years prior to full vacancy
decontrol provide one example.
However, restoration of vacancy control could not roll back current rent levels. Rent controls
must take into account the legitimate investment expectations of investors in rental property, who
in many cases have only recently purchased the property and paid much of the profit from
increased land values to the previous owners. That is why rent controls must start with the
current or a recent year as the base year, rather than rolling back rents to a previous level. It
would take at least a generation for stronger rent controls to roll back land rents in any Bay Area
city to a significant degree. Furthermore, tenants have no constitutionally protected rights to the
benefits of rent control and strong rent controls are vulnerable to the political process. No strong
rent control system other than New York City’s has survived over an extended period.
It remains an open question what can and should be done to reduce the excessive rent burdens of
most low-income tenants in Berkeley and the Bay Area.
46
Recommendation
The Rent Board should sponsor or cosponsor an affordable housing workshop or conference to
examine a range of approaches to improving maintenance, energy efficiency and seismic safety,
and making housing more affordable to low-income tenants. This should include consideration
of ways to generate additional funding for housing that is under alternative forms of ownership
such as land trusts, limited-equity cooperatives and non-profit housing corporations.
47
Section 7: Recommendations
I. The City should investigate measures to reduce the adverse impacts of the speculative
element in investment in rental property in Berkeley.
Such measures could include requirements that take effect when a rental property is
transferred or refinanced, such as:
Mandatory code inspection prior to or no more than 30 days after the time of sale or
refinance and all code violations cleared.
Energy efficiency inspection prior to or within 30 days of the time of sale or refinance
and must meet required standards (possibly as set by RECO, CECO).
Soft story buildings fully retrofit to meet a life-safety standard within one year of
transfer.
II. The Rent Board should provide an expedited Individual Rent Adjustment process that
allows increases in the rent ceiling for properties where the aggregate AGAs for all units
in the property have not provided a rent adjustment that matches the increase in the Bay
Area CPI-All Items.
III. The Rent Board should sponsor or cosponsor an affordable housing workshop or
conference to examine a range of approaches to improving maintenance, energy
efficiency and seismic safety, and making housing more affordable to low-income
tenants. This should include consideration of ways to generate additional funding for
housing that is under alternative forms of ownership such as land trusts, limited-equity
cooperatives and non-profit housing corporations.
48
Appendices
1. AGAs: 1981 - 2009
2. Numerical tables for all Figures
3. Vacancy Rate Data: Findings and Analysis
4. Estimating Land Rent in the Bay Area Rental Housing Market
49
Appendix 1: Annual General Adjustments, 1981 - 2010
1981: 5% OR 6.2% if owner paid for space heating
1982: 9% plus, if the owner paid for gas and electricity, including space heating, the
following additional increases were authorized: $4 for studio; $7 for 1-bedroom; $9 for 2-
bedroom; $10 for 3-bedroom; $12 for 4-bedroom; $16 for house with 3+ bedrooms.
Rental property owners who deferred the 1982 rent increase are entitled to a banking
bonus of an additional 1% for each year of deferral.
1983: 4.75% OR 5% if owner paid for electricity or gas
1984: 0%
1985: 2%
1986: 3% plus $2.50 per month
1987: 3.5%
1988: $25.00 per month OR $15.00 per month for residential hotels
1989: 3% OR 3.5% if owner paid for gas, electricity, or heating within the unit
1990: $16.00 per month
1991: 4% or $17.00 per month, whichever is higher
1991: 1991 Inflation Adjustment Order--45% of the 5/31/80 rent
1992: $26.00 per month
1993: $20.00 per month
1994: $18.00 per month
1995: 1.5%
1996: 1%
1997: 1.15%
1998: 0.8%
1999: 1% not to exceed $8.00
2000: $6
2001 : $10, plus an additional $8 if the owner pays for all gas service to the rental unit
2002 : 3.5% or $30, whichever is less, plus an additional $9 if the owner pays for interior space
heating to the rental unit
2003 : 0%
2004 : 1%, plus an additional $3, for units with new tenancies between 1/1/99 and 12/31/02;
1.5%, plus an additional $3, for units with no new tenancies since 1/1/99.
2005 : 0.9% (9/10ths
of a percent)
2006 : 0.7% (7/10ths
of a percent)
2007 : 2.6%
2008 : 2.2%
2009: 2.7%
2010: 0.1% (1/10th
of a percent)
Beginning with the 2000 AGA, the AGA is not applicable to units with new tenancies
starting in the previous calendar year. Example: the year 2000 AGA does not apply for
units with qualifying new tenancies during 1999.
50
Appendix 2: Numerical Tables for Figures
This appendix provides the numerical data that provided the basis for figures that did not include
the numbers on or below the figure itself. Figure 1: Bay Area & US Real Rent, 1935 - 2008
Real Rent (CPI-R/CPI-LS) 1935-2008 Year
US Real Rent
Bay Area Real Rent
Ratio of Change in Bay Area vs. U.S. Real Rent
1935 100.0 100.0 100.0
1936 100.90 99.2 98.4
1937 102.61 99.6 97.1
1938 109.19 105.0 96.2
1939 110.75 107.1 96.7
1940 109.96 106.4 96.7
1941 106.25 101.3 95.3
1942 96.18 91.5 95.1
1943 89.28 85.9 96.2
1944 88.18 85.5 96.9
1945 85.84 83.7 97.5
1946 78.63 76.0 96.7
1947 70.04 67.9 96.9
1948 69.19 66.4 96.0
1949 73.34 68.9 93.9
1950 75.60 69.3 91.7
1951 72.46 69.1 95.3
1952 73.91 70.8 95.8
1953 77.82 75.0 96.3
1954 80.57 77.5 96.2
1955 82.01 79.4 96.8
1956 82.46 80.2 97.2
1957 81.50 80.2 98.4
1958 80.39 80.8 100.5
1959 81.14 82.9 102.2
1960 80.89 84.5 104.5
1961 80.90 86.0 106.3
1962 81.16 88.1 108.5
1963 80.97 89.5 110.5
1964 80.78 90.6 112.2
1965 80.35 91.2 113.5
1966 79.15 90.6 114.5
1967 78.43 90.9 115.9
1968 77.19 91.9 119.1
1969 76.15 92.8 121.9
1970 75.49 94.7 125.5
51
1971 75.86 95.7 126.1
1972 76.15 95.5 125.5
1973 74.34 92.3 124.1
1974 70.26 86.5 123.1
1975 67.76 82.8 122.2
1976 67.41 82.8 122.8
1977 67.18 82.9 123.3
1978 67.26 82.7 123.0
1979 65.51 80.4 122.8
1980 63.84 80.1 125.5
1981 63.19 81.4 128.9
1982 64.33 83.0 129.0
1983 65.62 89.0 135.6
1984 66.30 91.7 138.3
1985 68.36 96.4 141.0
1986 71.66 103.2 144.0
1987 72.16 105.0 145.6
1988 72.14 104.8 145.3
1989 71.45 103.8 145.3
1990 70.63 103.8 147.0
1991 70.22 103.2 146.9
1992 69.99 101.9 145.5
1993 69.54 101.7 146.2
1994 69.58 102.1 146.7
1995 69.47 101.5 146.1
1996 69.36 102.2 147.4
1997 69.95 106.2 151.8
1998 71.62 113.2 158.1
1999 72.49 118.0 162.8
2000 72.61 122.4 168.5
2001 74.06 132.1 178.3
2002 76.49 136.4 178.3
2003 77.00 133.5 173.4
2004 76.99 129.8 168.7
2005 76.39 126.7 165.8
2006 76.74 123.9 161.4
2007 78.08 124.5 159.5
2008 77.46 124.8 161.2
52
Figure 2: Median Berkeley Rents, 1978 - 2009
Median Rent plus AGA
Median After Vacancy Decontrol
Median Vacancy Rent, 1 BR Year
$212
1978
$212
1979
$223
1980
$234
1981
$255
1982
$267
1983
$267
1984
$273
1985
$283
1986
$293
1987
$318
1988
$328
1989
$344
1990
$361
1991
$461
1991
$487
1992
$507
1993
$525
1994
$533
1995
$538 $570
1996
$545 $606
1997
$549 $643
1998
$554 $690 $950 1999
$560 $750 $1,100 2000
$570 $810 $1,200 2001
$590 $882 $1,150 2002
$590 $900 $1,100 2003
$602 $932 $1,050 2004
$608 $950 $1,095 2005
$612 $994 $1,100 2006
$628 $1,042 $1,200 2007
$642 $1,100 $1,275 2008
$659 $1,145 $1,295 2009
Note: Rent Board records on median rents are not available on an annual basis prior to 1998. Median rents for 1996-
97 in the “Median After Vacancy Decontrol” column are imputed on a straight line basis to represent the partial
vacancy decontrol increases allowed during the transition period. This is not entirely accurate, since part of the
difference between the “Median plus AGA” column and the “Median After Vacancy Decontrol” column is actually
IRAs from previous years.
53
Figures 3 & 4: Change in Berkeley Rent and Bay Area CPI and CPI-Rent 1978 - 2009
Year Median Rent plus AGA
Median Rent After Vacancy Decontrol
Median Vacancy Rent, 1 BR CPI
CPI-LS CPI-R
1978 100
100 100 100
1979 100
109 110 107
1980 105
125 125 121
1981 110
141 135 133
1982 120
152 145 146
1983 126
153 149 160
1984 126
162 157 174
1985 128
169 161 188
1986 133
174 163 204
1987 138
179 168 213
1988 150
187 175 222
1989 154
197 184 231
1990 162
205 192 242
1991 170
214 201 250
1991 217
214 201 250
1992 229
222 208 256
1993 239
228 214 263
1994 247
231 217 268
1995 251 251
236 222 272
1996 254 268
241 226 279
1997 256 285
249 231 296
1998 258 302
257 233 319
1999 261 324 460 268 240 342
2000 264 352 532 280 247 366
2001 269 380 581 295 253 405
2002 278 414 557 300 255 420
2003 278 423 532 305 260 420
2004 284 438 508 309 267 420
2005 286 446 530 315 275 421
2006 288 467 532 325 285 427
2007 296 489 581 336 295 444
2008 302 517 617 346 306 462
2009 310 538 627 348 305 477
Note: 2009 CPI is for the month of June.
54
Figure 5: Market & Stabilized Rents, 1998 - 2008, 1 BR Apartment
Year
1998
Tenant
Vacancy
Rent
1999
Tenant
2000
Tenant
2001
Tenant
2002
Tenant
2003
Tenant
2004
Tenant
2005
Tenant
2006
Tenant
1998 $624
1999 $630 $950 $950
2000 $636 $1,100 $950 $1,100
2001 $646 $1,200 $960 $1,100 $1,200
2002 $669 $1,150 $990 $1,130 $1,200 $1,150
2003 $669 $1,100 $990 $1,130 $1,200 $1,150 $1,100
2004 $682 $1,050 $1,003 $1,144 $1,215 $1,165 $1,100 $1,050
2005 $688 $1,095 $1,012 $1,155 $1,226 $1,175 $1,110 $1,050 $1,095
2006 $693 $1,100 $1,019 $1,163 $1,235 $1,183 $1,118 $1,057 $1,095 $1,100
2007 $711 $1,200 $1,046 $1,193 $1,267 $1,214 $1,147 $1,085 $1,123 $1,100
2008 $726 $1,275 $1,069 $1,219 $1,294 $1,241 $1,172 $1,109 $1,148 $1,124
55
Figure 6: Average Sales Price Per Unit, Berkeley Rental Properties with 5+ Units
Built Prior to 1980.
Year Price Per Unit
1978 PPU + CPI
Properties Sold
Units Sold
1977 $16,154 $16,154 8 110
1978 $23,593 $23,593 14 167
1979 $12,759 $26,271 8 141
1980 $25,404 $29,817 6 52
1981 $18,600 $32,892 1 5
1982 $18,986 $34,919 3 69
1983 $26,657 $36,041 4 67
1984 $24,468 $37,597 15 204
1985 $44,549 $38,935 8 61
1986 $38,176 $39,659 20 176
1987 $28,277 $41,106 18 159
1988 $32,253 $42,807 20 182
1989 $31,476 $44,870 17 172
1990 $47,965 $47,294 10 71
1991 $51,813 $49,284 11 75
1992 $47,545 $50,768 13 171
1993 $31,611 $52,288 19 312
1994 $48,527 $53,627 25 354
1995 $54,222 $55,146 23 232
1996 $51,640 $56,775 40 352
1997 $46,973 $58,077 23 253
1998 $52,962 $58,982 36 354
1999 $81,377 $60,285 28 326
2000 $94,436 $62,311 23 243
2001 $109,086 $64,084 22 203
2002 $106,858 $65,097 18 166
2003 $126,415 $66,581 31 331
2004 $139,033 $68,354 45 455
2005 $149,728 $70,670 57 517
2006 $134,314 $72,950 20 253
2007 $160,927 $75,027 34 344
2008 $166,486 $77,908 23 276
2009 $162,486 $81,198 15 158
56
Figure 7: Change in Berkeley Rents and 5+ Unit Property
Values, 1978 – 2009
RealQuest Alameda County data, CPI, Rent Board
Year AGA
Change in
Median Rent
Change in 1 BR
Vacancy Rent CPI
Change in value of 5+ unit
properties
1978 100 100
100 100.0
1979 100 100
109 60.2
1980 105 105
125 128.8
1981 110 110
141 63.4
1982 120 120
152 90.4
1983 126 126
153 106.9
1984 126 126
162 118.3
1985 128 128
169 221.0
1986 133 133
174 166.3
1987 138 138
179 144.9
1988 150 150
187 132.4
1989 154 154
197 163.4
1990 162 162
205 200.0
1991 170 170
214 239.6
1991 217 217
214 239.6
1992 229 229
222 228.3
1993 239 239
228 177.0
1994 247 247
231 219.3
1995 251 251
236 244.6
1996 254 268
241 257.6
1997 256 285
249 220.8
1998 258 302
257 249.1
1999 261 324 460 268 345.2
2000 264 352 532 280 410.9
2001 269 380 581 295 481.0
2002 278 414 557 300 522.4
2003 278 423 532 305 598.3
2004 284 438 508 309 630.4
2005 286 446 530 315 656.5
2006 288 467 532 325 577.5
2007 296 489 581 336 656.9
2008 302 517 617 346 729.7
2009 310 538 627 351 610.0
57
Figure 9: Berkeley and Alameda County
price per square foot, 1990 – 2009
CoStar data
Year
CoStar Berkeley 5+ units, median
CoStar Alameda County, 5+ units, median
1990 $52.69 $68.64
1991 $51.76 $59.22
1992 $52.91 $57.93
1993 $53.38 $50.72
1994 $59.10 $56.73
1995 $60.10 $55.84
1996 $61.88 $57.58
1997 $76.71 $67.93
1998 $80.91 $64.66
1999 $95.92 $67.96
2000 $120.41 $79.28
2001 $125.63 $102.35
2002 $128.93 $113.64
2003 $166.41 $131.86
2004 $176.06 $142.06
2005 $179.11 $163.23
2006 $176.27 $154.02
2007 $192.94 $173.04
2008 $228.75 $165.97
2009 $159.00 $148.65
58
Figure 10: Average Berkeley price per square foot, Alameda County property records via RealQuest
Year 5+ unit properties Four Unit Properties
Two unit properties
1978 $31.49 $26.89 $42.62
1979 $18.96 $37.04 $58.84
1980 $40.55 $26.69 $55.61
1981 $19.96 $36.83 $59.33
1982 $28.46 $35.76 $52.93
1983 $33.66 $33.46 $61.06
1984 $37.24 $46.85 $61.69
1985 $69.60 $47.73 $78.85
1986 $52.36 $61.27 $72.14
1987 $45.62 $59.18 $89.47
1988 $41.68 $63.77 $110.51
1989 $51.46 $58.12 $124.90
1990 $62.98 $87.11 $122.88
1991 $75.44 $80.85 $108.91
1992 $71.90 $63.50 $120.83
1993 $55.75 $54.33 $131.49
1994 $69.04 $84.64 $154.64
1995 $77.01 $87.61 $127.59
1996 $81.11 $71.96 $119.64
1997 $69.54 $82.00 $140.53
1998 $78.44 $103.12 $155.74
1999 $108.69 $132.02 $159.28
2000 $129.39 $150.45 $196.55
2001 $151.45 $174.09 $236.64
2002 $164.49 $194.91 $257.75
2003 $188.40 $201.07 $302.96
2004 $198.50 $247.07 $341.67
2005 $206.74 $281.55 $382.16
2006 $181.84 $294.90 $387.41
2007 $206.84 $284.26 $385.01
2008 $229.78 $262.77 $368.50
2009 $192.07 $203.98 $283.13
59
Table for Figure 13: 2007 Median Rental Income and Expenses Per Square Foot
for Garden Apartment Buildings in Selected Metropolitan Areas
Income
Operating
Expenses
OE as %
Income NOI
NOI as %
Income
U.S. $10.44 $4.97 48% $5.51 53%
Houston, TX $9.74 $5.63 58% $4.14 43%
Portland, OR $9.76 $4.30 44% $5.81 60%
Minneapolis, MN $10.00 $5.14 51% $4.67 47%
Sacramento $12.28 $5.45 44% $6.50 53%
Oakland $18.14 $6.99 39% $11.23 62%
San Francisco $18.63 $6.07 33% $11.91 64%
Los Angeles $20.13 $6.69 33% $13.45 67%
San Jose $20.63 $6.89 33% $14.24 69% Source: IREM Institute of Real Estate Management, "Income/Expense Analysis: Conventional
Apartments", 2008, and author’s calculations. Due to use of medians, expenses and NOI will not
add up exactly to rental income.
60
Appendix 3: Vacancy Rate Data: Findings and Analysis
Findings from the 2009 Tenant Survey
Postal returns from the 2009 tenant survey indicate that the vacancy rate in April 2009 was 4.0%
for units in the sampling universe – registered rental units and units with tenants in the Section 8
and Shelter Plus Care Programs. The month of April was chosen for the survey because it is the
month before students take their final examinations so that it has significantly less turnover than
the summer months.
The survey also found 38% of units occupied by tenants who moved in during the previous 16
months. Using an assumption that there is an average of one to two months between the time one
tenant leaves and another moves in, Berkeley’s rent stabilized units would have an annual
vacancy rate of from 2.4% - 4.8%.
In comparison, the most regular and readily available vacancy rate data is provided by the U.S.
Census Bureau’s American Community Survey (ACS) which estimated Berkeley’s overall rental
vacancy rate at 3.8% for the 2005-2007 period and 2.4% for the 2006-2008 period. This data is
for all rental units, including rented single-family houses and condominiums and rental housing
built after 1980. The Census Bureau does not make this data available for individual years for
Berkeley because sample size is too small to provide accurate results.
The Mail Survey and Postal Returns
In April 2009 the RSB sent a mail survey to “Tenant” at 1,809 addresses of registered rental
units and units temporarily exempt because the tenant was said to be in the Section 8 or Shelter
Plus Care rental assistance programs. Three separate mailings went to all survey units: an April
6th
introductory letter, an April 10th
letter containing the survey, and an April 17th
postcard
reminder. The Post Office returned at least one of these three mailings from 94 different
addresses. Of the 94 addresses with returns, 66 were returned with labels indicating that they
were vacant or likely vacant, including “vacant”, “unclaimed”, “unable to forward”, “box full”.
The other 28 were returned with labels such as “no such number”. These were not counted as
vacant but rather as undeliverable, either because of incorrect addresses in the Rent Board
database or postal error. There is no way to determine how many units were vacant where none
of the mailings were returned, nor is there any way to know how many of the apparent vacancy
returns were incorrectly labeled or returned. We received several completed questionnaires from
units from which the post office also returned one of the mailings and removed these from the
returned list.
As described previously, the survey went to two sampling groups, a one-in-fifteen sample and a
one-in-five sample. The one-in-fifteen sample represented 78% of all registered units and the
one- in-five sample represented 22%. Removing the undeliverable units from the base mailing
totals and assigning the vacancies proportionately for the two sample groups, we arrive at a
vacancy rate of 4.04% for the month of April.
In the 2009 tenant survey 38% of respondents had moved into their unit during the previous 16
months. This indicates a turnover rate of 28.5% annually. If each unit is vacant for an average of
61
one month, then the annualized vacancy rate would be 2.4%. If the average vacancy is two
months, then the annualized vacancy rate would be 4.8%.
The Policy Implications of Vacancy Rates
The real estate industry frequently argues that a rental vacancy rate of four or five percent means
that there is a “balanced” housing market and that this means that rent and condominium
conversion controls are unnecessary. This involves a fundamental misunderstanding of the role
of vacancies in the rental housing market, particularly in high-rent areas such as the coastal
regions of California. Vacancy rates are important to the housing market as a necessary
precondition for adequate tenant mobility and as an indicator of the balance between supply and
effective demand. However, vacancy rates often do not serve as a good indicator of the balance
between supply and demand because they are affected by cyclical short-term economic changes
as well as by the underlying balance between the supply and the need for housing. The
fundamental indicator of “balance” in a housing market is rent levels and how the rents compare
with other markets.
Vacancy rates go up and down with the state of the overall economy and they also go up and
down with the rate of development of new housing in relation to population growth and
household formation, so at any given time the vacancy rate reflects both of these factors.
Vacancy rates typically increase during recessions, which reduce what consumers are able to pay
for housing and thus reduce what economists call “effective demand” for housing. Such an
increase in vacancies does not mean that the need for housing has decreased, nor does it mean
that the utility of housing to consumers has decreased. Rather, it simply means that due to the
recession consumers have less money with which to purchase housing and landlords have not
reduced rents in step with reduced consumer ability to pay. Even if vacancies increase, landlords
will be reluctant to lower rents for new tenants, since they may then be pressured to reduce rents
for current tenants who moved in at higher rents. Within limits the loss of revenue from vacant
units may be less than the loss of revenue from reducing rents for tenants already in place.53
Only if a recession is so severe that it leads to a decline in population does the actual need for
housing also decline during such a short-term economic cycle.
The coastal areas of urban California have a long-term, underlying housing shortage, caused by
the widely recognized shortfall of new housing development in relation to population growth.54
Despite this it is perfectly normal to also have short-term increases in the vacancy rate that result
from a recession or economic slow-down. Then when the cycle changes, vacancies go down and
rents start to increase again.
53
Eric Belsky, “Rental Vacancy Rates: A Policy Primer”, Housing Policy Debate, V.3#3,
1992, pp.793-813,
http://www.fanniemaefoundation.org/programs/hpd/pdf/hpd_0303_belsky.pdf. 54
John Landis, Michael Smith-Heimer, et al. Raising the Roof: California Housing Development Projections
and Constraints, 1997-2020, Department of Housing and Community Development, Sacramento, CA 2000.
Hans Johnson, Rosa Moller & Michael Dardia, In Short Supply? Cycles and Trends in California Housing,
Public Policy Institute of California, San Francisco, March 2004.
62
In summary, Bay Area vacancy rates are subject to short-term cycles varying with the state of the
local economy; even though there is a long-term shortage of housing that is expected to last for
the foreseeable future. An increase in the vacancy rate in the Bay Area does not mean that
sufficient new housing has been constructed to balance the market, nor does it mean that there is
a sufficient housing supply to stabilize market prices over the long run without public
intervention. For this reason, it is important not to respond to short-term fluctuations in the
vacancy rate with changes in public policies that will result in a long-term loss of rental units.
Condominium conversion permanently removes the great majority of the converted units from
the rental market.
Finally, even looking at long-term vacancy rates, there is no consensus on what vacancy rate is
necessary to balance the market. Many ordinances use a five percent long-term vacancy rate as
the measure of a healthy rental market, largely because ordinances tend to follow one another.
The empirical evidence on appropriate vacancy rates is mixed. Gabriel and Nothaft support the
five percent estimate, while Gilderbloom and Appelbaum review a number of studies and
suggest that a vacancy rate of 9% or 10% is necessary.55
The U.S. Census Housing Vacancy
Survey has reported national metropolitan area rental vacancy rates averaging over 7% since
1990 and in the 8-10% range since 2001, even as rents continued to rise, further evidence that the
“normal” vacancy rate is generally higher than 5%. Eric Belsky argues that the equilibrium or
“natural” vacancy rate varies from one local market to the next, and will also vary by size of
building and property holdings, since small landlords tend to “minimize vacancies” while larger
landlords tend to “maximize rents”. 56
The table below shows median contract rents and vacancy rates as reported for the twenty-six
largest metropolitan areas by the Census Bureau’s 2006-2008 American Community Survey. It
shows that lower-rent areas tend to have higher vacancy rates. The majority of the areas with
rents in the bottom third have vacancy rates over 10%, most of the areas in the middle third have
vacancy rates over 8%, and all of the areas in the top third have vacancy rates under 8%. Even
so, there is a great deal of variation among regions with similar rents.
Vacancy rates are only one indicator of the state of the housing market. Prices are a much more
fundamental indicator. If housing costs in an area are well above average and remain so over a
long period of time, it means that there is an underlying housing shortage regardless of the
current vacancy rate.
55
Gabriel, Stuart A. & Frank E. Nothaft, “Rental Housing Markets, the Incidence and Duration of Vacancy
and the Natural Vacancy Rate”, Journal of Urban Economics, V.49, 2001, pp.121-149.
John I. Gilderbloom & Richard P. Appelbaum, Rethinking Rental Housing, Temple University Press,
Philadelphia, 1988, pp.52-56. 56
Eric Belsky, “Rental Vacancy Rates: A Policy Primer”, Housing Policy Debate, V.3#3, 1992, pp.793-813,
http://www.fanniemaefoundation.org/programs/hpd/pdf/hpd_0303_belsky.pdf
63
Rents and Vacancy Rates for the 25 Largest US Metropolitan Areas
& Berkeley, Oakland, Alameda County, 2006-2008
Median contract rent Vacancy rate
Pittsburgh, PA Metro Area $495 9.10%
Cincinnati-Middletown, OH-KY-IN Metro Area $549 12.40%
St. Louis, MO-IL Metro Area $553 7.90%
Detroit-Warren-Livonia, MI Metro Area $637 10.90%
Houston-Sugar Land-Baytown, TX Metro Area $638 12.50%
Dallas-Fort Worth-Arlington, TX Metro Area $665 11.20%
United States $676 7.90%
Portland-Vancouver-Beaverton, OR-WA Metro Area $720 5.40%
Atlanta-Sandy Springs-Marietta, GA Metro Area $738 12.00%
Tampa-St. Petersburg-Clearwater, FL Metro Area $741 10.40%
Denver-Aurora, CO Metro Area $743 7.80%
Philadelphia-Camden-Wilmington, PA-NJ-DE-MD Metro Area $748 8.70%
Minneapolis-St. Paul-Bloomington, MN-WI Metro Area $761 6.80%
Chicago-Naperville-Joliet, IL-IN-WI Metro Area $766 8.20%
Phoenix-Mesa-Scottsdale, AZ Metro Area $777 10.60%
Baltimore-Towson, MD Metro Area $812 8.40%
Seattle-Tacoma-Bellevue, WA Metro Area $839 5.10%
Sacramento--Arden-Arcade--Roseville, CA Metro Area $892 6.60%
Oakland city, California $912 8.60%
Miami-Fort Lauderdale-Pompano Beach, FL Metro Area $923 9.20%
Riverside-San Bernardino-Ontario, CA Metro Area $950 6.60%
New York-Northern New Jersey-Long Island, NY-NJ-PA Metro Area $956 4.60%
Boston-Cambridge-Quincy, MA-NH Metro Area $983 5.50%
Los Angeles-Long Beach-Santa Ana, CA Metro Area $1,046 3.80%
Berkeley city, California $1,069 2.40%
Alameda County, California $1,091 6.70%
San Diego-Carlsbad-San Marcos, CA Metro Area $1,126 4.90%
Washington-Arlington-Alexandria, DC-VA-MD-WV Metro Area $1,127 7.20%
San Jose-San Francisco-Oakland, CA CSA $1,188 5.20%
Source: U.S. Census Bureau, American Community Survey 2006-2008
Notes: Vacancy rate is “vacant for rent” divided by occupied rentals plus vacant for rent. Geographical grouping is
Standard Metropolitan Statistical Areas except that the San Francisco Bay Area is the eleven county Consolidated
Statistical Area, which includes six MSAs: San Francisco-Oakland, San Jose, Santa Rosa, Vallejo, Napa and Santa
Cruz-Watsonville.
64
Appendix 4: Estimating Land Rent in the Bay Area Rental Housing Market
Land Rent
The people who live and work together in urban areas collectively generate economic, cultural
and social benefits. This is partly done through government, the institution we use to provide
public safety, transportation, education and other systems that sustain urban life. It is partly
through the simple fact that urban areas generate dense networks of human interaction that
advance knowledge and creativity in every field and endeavor, whether it is the arts, the sciences,
business, government or ways of life. Locations within these areas with high concentrations of
desirable human interactions, activities and services become valuable because available space is
limited. More housing can be produced on the existing centrally located land only by building at
higher densities, which increases the cost of producing the housing. More land to build on is
available only on the outskirts of the area, away from the desirable activities.
Residential real estate is a form of property that combines buildings and the land the buildings sit
on. When people buy a home, part of what they pay is for the building and part is for the land it
sits on. When people rent an apartment part of the rent payment supports construction, operation
and maintenance of the building (building rent) and part of the payment is for access to that
location (land rent). The value of this location comes from a combination of public investment
and the network of activities generated by people who live and work in the surrounding area.
There is nothing new in these observations about the rental housing market. Adam Smith, often
considered the founding figure in economics, wrote about it in his book The Wealth of Nations,
published in 1776.
“The rent of a house may be distinguished into two parts, of which the one may very properly be
called the Building rent; the other is commonly called the Ground rent. … The building rent is
the interest or profit of the capital expended in building the house …and secondly, to keep the
house in constant repair. …Whatever part of the whole rent of a house is over and above what is
sufficient for affording this reasonable profit, naturally goes to the ground-rent… revenue which
the owner, in many cases, enjoys without any care or attention of his own”57
In the same pages, Smith also argued that land value or land rent is value created by the public,
saying “Ground-rents, so far as they exceed the ordinary rent of land (its agricultural value), are
altogether owing to … good government...” He further argued that it would be particularly
appropriate to tax land rent, since “A tax upon ground-rents would not raise the rents of houses.
It would fall altogether upon the owner of the ground rent…”
The investors who own the residential land in successful urban areas profit from its increased
value even when they have contributed little or nothing to that increase. This disconnect between
creation of land value and profit from land value results in a cruel irony for many residents who
57
Adam Smith, An Inquiry into the Nature and Causes of the Wealth of Nations, 1776, Book V, Chapter 11, Part
2. (Modern Library Edition, 2000, pages 904-5, 909.)
65
do not own real estate. They contribute to making the city they live in a better and more
interesting place and in so doing they increase land values, which increases the rent they have to
pay to continue to live there. A well-known example of this is the common pattern in which
artists who live in a low-rent neighborhood are then forced out by rising rents when it is
“discovered” by higher-income people.
When people cannot afford to pay enough rent to cover the costs of operating and maintaining a
decent apartment building they have an income problem. But when people have enough money
to cover those costs and still can't find decent quality housing they can afford, then housing
affordability is a problem of land rent.
Land Rent in the San Francisco Bay Area Rental Housing Market
The San Francisco Bay Area has the highest rents of any major metropolitan area in the U.S.
Figure 1 shows the 2008 median rents for the U.S. and the twenty-five largest metropolitan areas
as reported in the U.S. Census Bureau’s American Community Survey for 2006-2008. The
median rent in all U.S. metropolitan areas is 43% lower than the Bay Area rent. The median for
all U.S. cities is near the lower end of this table because the U.S. average includes all the smaller
metropolitan areas, which tend to have lower densities and housing prices. Even the 25th
percentile rent in the Bay Area is $882, higher than the median rent in most other major
metropolitan areas. (Table 1 in the appendix shows rents by quartile.)
66
The sources of land rent in the Bay Area rental housing market are a combination of the factors
that make the Bay Area a desirable place to live and the factors that severely constrain the
availability of land that can be developed with multifamily housing. The Bay Area has a strong
economy led by Silicon Valley and biotechnology firms; high-quality infrastructure created by
public investment in world-class universities, highways, mass transit and parks; a beautiful
environment featuring the Bay and the California coast; and an open, diverse regional culture
that is highly attractive to creative people. At the same time, it features extraordinary barriers to
housing development, including difficult geography, with a coast, a bay and hilly areas with
steep slopes; highly developed local government land use regulations; and a central area that is
already relatively dense by U.S. standards, which makes increases in density more costly because
they often require redevelopment of sites already built on and more costly forms of high-rise
construction.
There are excellent overviews of the role of land values in the single-family housing market.
Davis and Palumbo used data on single-family home prices to separate out the value of buildings
and land in the homeownership market in the 46 largest metropolitan areas of the United States.58
They found that in 1984 in the largest 46 metropolitan areas the land value averaged 32% of the
value of a single-family home, a figure that ranged from 11% in the Midwest to 55% on the West
Coast, with the value in land reaching 61% in the Oakland area and 75% in the San Francisco
area. By 2004, near the high point of the 1998-2007 “housing bubble”, a nationwide average of
51% of the value of a single-family home in the 46 largest metropolitan areas could be attributed
to the land and the proportion ranged from 36% of total value in the Midwest to 74% in the West
Coast states. In the Oakland area 78% and in the San Francisco area 89% of the value of the
average single-family home was in the land by 2004, the highest ratio among the major
metropolitan areas. While the collapse of the housing bubble may return the U.S. to something
closer to the 1984 price structure, land values clearly constitute a major component of ownership
housing prices in many metropolitan areas, and are particularly high the San Francisco Bay Area.
Similar studies of rental housing are not available and far less data is available on rental housing,
but there are readily available data sources to produce rough estimates of the magnitude of land
rent in the Bay Area rental housing market.
There are, however, three possible alternative explanations for the Bay Area’s higher rents:
higher housing quality, higher operating costs, higher construction costs. I will examine each of
these possibilities in turn and provide estimates of the extent to which each of these may explain
the Bay Area’s higher rents. The part that cannot be explained by differences in quality and cost
is the land rent.
58
Davis, Morris A. and Michael G. Palumbo. (2006). “The Price of Residential Land in Large U.S. Cities”, Federal
Reserve Board Finance and Economics Discussion Series, No. 2006-25.
67
Alternative Explanation: Housing Quality
There is an excellent source of data on residential rents that corrects for changes in quality – the
Bureau of Labor Statistics' Consumer Price Index residential rent component. The Consumer
Price Index (CPI) takes changes in quality into account, so that CPI increases for particular
components of consumer spending, such as rents, reflect actual price increases for goods of the
same quality. CPI data is available nationally, for all U.S. cities, and separately for the largest
metropolitan areas, including the San Francisco Bay Area. The CPI rent component is adjusted
for changes in the size and amenities of the units in the sample, which is periodically updated to
take into account changes in the composition of the housing stock. This CPI data allows
comparisons between changes in rents over time in the Bay Area and in other metropolitan areas
as well as in the average change for all U.S. Cities.59
Figure 2 shows the trend in Bay Area and US rents measured against the change in the CPI-All
Items Less Shelter since 1935, when the Less Shelter index first becomes available. U.S. and
Bay Area real rents follow each other closely until the late 1950s. After that point U.S. rents
remain largely stable, while Bay Area rents increase. Taking the ten year post-war period of
1946 to 1955 as a base-line, by 2008 Bay Area rents were 69% higher than they would have been
if they had changed in the same way as U.S. rents over all. Put the other way, if Bay Area rents
had followed the trend of rents in U.S. cities overall, they would have been 41% lower and they
would be comparable to rents in the Portland, Oregon area.60
59
Use of the CPI-All Items to measure increases in rent in the Bay Area would run into a problem of circularity.
When rents are going up faster than the cost of other consumer goods, as they have in the Bay Area, this increases
the CPI-All Items. Using a measure that includes rent increases to measure the rent increases will mask part of the
increase. The Bureau of Labor Statistics CPI index for all items except Shelter allows us to measure changes in rents
without this source of distortion.
The CPI-All Items Less Shelter index is available from 1935 to the present for all US cities, but only from
1976 to the present for the San Francisco Bay Area, so for changes in Bay Area rents in relation to CPI-LS I used a
omposite of the national index from 1935-1976 and the annual change in the Bay Area index from 1976-2008. The
CPI-All Items Less Shelter increased at virtually the same rate in the Bay Area and in all U.S. cities between 1976
and 2008, so use of the composite index makes little difference. This suggests that the Bay Area’s larger increase in
the CPI-All Items is mostly the result of the Bay Area’s more rapidly increasing housing costs. 60
There is reason to believe that the historical CPI data underestimates changes in rent, so that the quality-adjusted
gap between the Bay Area and the rest of the U.S. is even greater than it appears. Between 1953 and 1994 the
Bureau of Labor Statistics made a number of improvements in the way it gathers data on rents and these corrections
are retroactively estimated in Crone, Theodore, Leonard Nakamura, Richard Voith, “Rents Have Been Rising, Not
Falling, in the Post-War Period”, Working Paper 08-28, Federal Reserve Bank of Philadelphia, 2008.
68
Figure 3 compares changes in real rents since 1935 for the Bay Area and Los Angeles area and
several major metropolitan areas with more balanced housing markets: Portland, Oregon;
Minneapolis-St. Paul; Houston, Texas. The trend for Los Angeles has been somewhat different
from the Bay Area’s, but they have diverged nearly as much from the trend for the U.S. as a
whole.
Portland is a successful, growing west coast metropolitan area well known for the high quality of
its urban life. Minneapolis-St. Paul is a Midwestern area also known for its high quality of life.
These metropolitan areas, along with San Francisco, San Jose and Oakland, are among the top
ten in Prof. Richard Florida’s urban “creativity index”, metropolitan areas that score highly on
“technology, talent, and tolerance”.61
Florida comments that “The greater Minneapolis region
combines a strong creative economy with low rates of poverty, affordable housing and a
balanced income distribution.”62
In the Houston area land values and land rents are at a
minimum. Houston is widely known as the only major city in America without zoning and
weaker land use regulations.63
61
Florida, Richard. (2005). The Flight of the Creative Class, HarperCollins, N.Y., pp.283-4. 62
Ibid, p.262. 63
Saiz, Albert. (2008). “On Local Housing Supply Elasticity”, SSRN: http://ssrn.com/abstract=1193422.
Siegan, Bernard. (1972). Land Use Without Zoning, Lexington.
69
The comparison between the Bay Area and the U.S. as a whole is a conservative way to estimate
land rent. There is an element of land rent in virtually every local housing market, so any
comparison of the Bay Area with other cities will simply show how much more land rent is to be
found in the Bay Area rather than the total amount of land rent. For example, even in the
Portland, Oregon area, with a much more balanced housing market and much lower rents, a
study found that between1992 to 2002 “increased population, coupled with an essentially fixed
supply of land” caused real increases in apartment rents in the center and at major transportation
nodes, and “resulted in a wealth transfer from renters to owners”.64
The comparison with
Houston also suggests that U.S. rents already include significant land rent.
64
Wilson, Beth and James Frew. (2007). “Apartment Rents and Locations in Portland, Oregon: 1992-2002”, Journal
of Real Estate Research, V.29, p.214.
70
Alternative Explanation: Operating Expenses
We can look at whether the Bay Area’s higher rents can be explained by higher operating and
maintenance expenses by using data from the Institute for Real Estate Management (IREM). The
Institute publishes an annual “Income/Expense Analysis” with data on operating expenses, rent
collections and net operating income from a survey of its membership broken down by
metropolitan areas so that we can make the necessary comparisons.
Figure 4 shows graphically the median rental income, operating expenses and net operating
income per square foot for the U.S. and Bay Area and other selected cities as reported to IREM
for garden apartment buildings for 2007.65
(Garden apartment buildings are low-rise rentals with
associated green space, the largest group in the IREM sample and the only type that can provide
a comparison with all areas.) The comparison cities are those shown previously in examining
changes in the CPI rent component.
Median rents for garden apartments in the central Bay Area, which includes the metropolitan
areas of Oakland (Alameda and Contra Costa Counties), San Jose (Santa Clara County) and San
Francisco (Marin, San Francisco and San Mateo Counties) are from 74% to 98% higher than the
median rents nationwide. The variation in net operating income is substantially greater than the
65
IREM Institute of Real Estate Management, "Income/Expense Analysis: Conventional Apartments", 2008. Debt
service is an investment expense, not an operating expense, and is paid out of NOI.
71
variation in operating expenses. The median operating expenses in the Bay Area are from 22% to
40% above the U.S. median, while the NOI in the Bay Area is from 104% to 158% higher than
the U.S. median.
The higher expenses in the Bay Area average out to 9% of the average rent as do the higher
expenses in the Los Angeles area. (See Table 2)
Table 2: 2007 Median Rental Income and Expenses Per Square Foot for
Garden Apartment Buildings in Selected Metropolitan Areas
Income
Operating
Expenses
OE as %
Income NOI
NOI as %
Income
U.S. $10.44 $4.97 48% $5.51 53%
Houston, TX $9.74 $5.63 58% $4.14 43%
Portland, OR $9.76 $4.30 44% $5.81 60%
Minneapolis, MN $10.00 $5.14 51% $4.67 47%
Sacramento $12.28 $5.45 44% $6.50 53%
Oakland $18.14 $6.99 39% $11.23 62%
San Francisco $18.63 $6.07 33% $11.91 64%
Los Angeles $20.13 $6.69 33% $13.45 67%
San Jose $20.63 $6.89 33% $14.24 69%
Source: IREM Institute of Real Estate Management, "Income/Expense Analysis: Conventional Apartments", 2008.
The 2007 IREM data is based on 3,100 units in the Oakland area, 2,297 in the S.F. area, 3,782 in the San Jose area.
Due to use of medians, expenses and NOI will not add up exactly to rental income.
The higher NOI in the Bay Area means a typical rental property will have more than twice the
value of similar rental properties in many other parts of the U.S. With higher property values,
even with California’s property tax limitation, over half of the difference between average U.S.
operating expenses and the Bay Area and L.A. area is the result of higher property taxes. (See
Table 3) In effect, this is a small tax on land rent in the Bay Area and considering it as such
reduces the adjustment for higher operating expenses to 4% of the Bay Area rent. If we include
the public services paid for through real estate taxes as an essential aspect of housing, then the
higher operating cost differential is the appropriate comparison. Higher operating expenses thus
appear to explain from 4% to 9% of the 40% gap between Bay Area and U.S. rents.
72
Table 3: U.S. and Bay Area Operating Expenses Per Square Foot
Income
Operating
Expenses OE-US OE
Difference
as % of
Income
Real Estate
Taxes
U.S. $10.44 $4.97 $0.00 0.00% $0.89
Houston, TX $9.74 $5.63 $0.66 6.78% $1.03
Portland, OR $9.76 $4.30 -$0.67 -6.86% $0.82
Minneapolis, MN $10.00 $5.14 $0.17 1.70% $0.89
Sacramento $12.28 $5.45 $0.48 3.91% $0.85
Los Angeles $20.13 $6.69 $1.72 8.54% $1.54
Oakland $18.14 $6.99 $2.02 11.14% $1.46
San Francisco $18.63 $6.07 $1.10 5.90% $2.31
San Jose $20.63 $6.89 $1.92 9.31% $1.76 Source: IREM Institute of Real Estate Management, "Income/Expense Analysis: Conventional Apartments", 2008.
Alternative Explanation: Construction Costs
The San Francisco Bay Area has among the highest construction costs in the United States, but
this has been true for quite a long time. In order to compare changes in Bay Area construction
costs with other cities, we can refer to ENR, the former Engineering News Record, which
maintains a general purpose building cost index based on the cost of a fixed quantity of skilled
and unskilled labor and materials from 22 different urban areas.66
Figure 5 shows the change in the ENR building cost index from 1948 to 2008 adjusted by the
change in the CPI-Less Shelter. It shows that both U.S. and Bay Area building costs increased
rapidly from 1948 to 1972 and then flattened out for a decade and have declined since 1979. Bay
Area building costs increased faster than U.S. costs from 1962 to 1976, but have actually
declined slightly more than U.S. costs since then. On average, since 1962 the Bay Area’s real
costs have increased by 9.5% more than U.S. real costs. The 2008 American Community Survey
indicates that 60% of Bay Area rental housing was built after 1960 and construction costs are
paid for out of the net operating income. A 9.5% increase in NOI would increase rents by the
proportion of NOI, which often reaches two-thirds of the rent in newly constructed buildings, for
an increase in rent of 6.4% applied to the 60% of buildings built since 1962, which suggests that
higher construction costs potentially added up to 3.8% to Bay Area rents.
66
McGraw-Hill, ENR.com http://enr.construction.com. The cost index is for all forms of construction, not
specifically for residential construction.
73
However, while housing is a good that requires a major initial capital investment, the subsequent
cost of operation, maintenance and periodic renovations is much lower. Indeed, if newer
construction is higher quality it may be less costly to operate and maintain over the life of the
building, with lower utility costs due to better insulation for example. If there is sufficient
continuing production of new housing, which is usually directed towards higher income tenants,
then as it ages there will be a continuous stream of additional older housing that will compete for
tenants with existing older housing and this competition will reduce rents to closer to the actual
ongoing costs. Since construction costs are amortized over time, a reasonable argument can be
made that the effects of higher construction costs should only apply to buildings still in their
amortization period. The standard depreciation period of 27.5 years would take us back to 1980.
The American Community Survey (2006-8) indicates that 27% of Bay Area rental housing was
built since 1980, which would mean that higher construction costs are responsible for only 1.7%
of Bay Area rents.
74
Estimate of Land Rent in the Bay Area
Drawing together this analysis of data on rents adjusted by quality, operating costs and
construction costs, we arrive at the following rough estimate of the proportion of land rent in the
Bay Area rental housing market.
Difference between Bay Area and U.S. median rent, 2008 43%
Quality adjusted gap between Bay Area and US rents: 41%
Based on change in CPI-R/CPI-LS, 1946/1955 to 2008
Correction for higher operating costs: -9%
Operating costs excepting real estate taxes: -4%
Based on IREM 2007 data
Correction for higher construction costs: -2%
Based on ENR index, 1948-2008
Estimated land rent: 30% - 35%
The estimate above is necessarily very approximate given the limitations of the data used. It
serves to give a general sense of how much tenants in the Bay Area pay in pure land rent, over
and above the rent actually necessary to support their rental housing.
According to the Bureau of the Census American Community Survey for 2006-8 there were
1,064,000 tenant households in the eleven-county San Jose-San Francisco-Oakland Consolidated
Metropolitan Statistical Area and together they paid an aggregate monthly contract rent of $1.3
billion, for a total of $15.6 billion annually. If the land rent is 30% to 35% of the current rent
then it amounts to $4.5 to $5.5 billion annually in the Bay Area. In sum, Bay Area tenants are
paying around $5 billion more annually than would actually be necessary in order to profitably
operate and maintain the housing they live in. Partially as a result, one-third of Bay Area tenants
(340,000 households) pay more than 40% of their income in gross rent and one-quarter (250,000
households) pay over half of their income. These tenants with high rent burdens are mostly very-
low-income tenants with incomes below 50% of the area median. It seems clear that Bay Area
rents are limited less by competition and more by the limits of what tenants can manage to pay.67
67
As Adam Smith explained in An Inquiry into the Nature and Causes of the Wealth of Nations, (1776) “The price
of monopoly is upon every occasion the highest which can be got. The natural price, or the price of free competition,
on the contrary, is the lowest which can be taken, not upon every occasion indeed, but for any considerable time
together. The one is upon every occasion the highest which can be squeezed out of the buyers, or which, it is
supposed, they will consent to give: The other is the lowest which the sellers can commonly afford to take, and at
the same time continue their business.” Smith used the term “monopoly” broadly to include any situation where
competition is limited, not just for situations where there is only one seller. Book V, Chapter 11, Part 2. (Modern
Library Edition, 2000, page 69.)
75
Land Rent Is a Permanent Feature of the Bay Area Housing Market
The standard economists' response to high housing costs is to call for the elimination of
regulatory barriers to new construction. It is certainly true that reduction of regulatory barriers to
housing development at higher densities will increase supply and lower land values and land
rents to some degree. However this is unlikely to greatly reduce land rent in the Bay Area. High
rents and land values resulting from central urban locations were characteristic of successful
cities long before the development of modern land use regulations.
Not only is much of the central Bay Area is already quite dense, but there are major
environmental constraints on where housing can be built, most notably the bay itself. There was
a serious proposal in the 1950s to fill in most of the San Francisco Bay for development. This
would have added hundreds of square miles of new land in the heart of the Bay Area and further
reduced the value of the surrounding land by removing the Bay as an aesthetic and recreational
amenity. It was precisely this dystopian vision that lead in the 1960s to creation of the Save the
Bay movement and then establishment of the Bay Conservation and Development Commission
to protect the Bay and regulate shoreline development.
Moving farther out from the central area along the Bay, many suburban communities have
extensive restrictions on development of multi-family housing even though such development
would not have harmful effects on the regional environment. Indeed, a growth cap established by
the City of Pleasanton was recently overturned by the courts on the grounds that it was contrary
to state law requiring each city to provide for its “fair share” of projected growth. However,
additional construction of rental housing in outlying areas would have limited effects on the
urban core areas and even then only with a substantial delay. The effects of reductions in
regulatory barriers have been marginal at best. As shown in Figure 2, real rents began to increase
faster than those in the U.S. as a whole back in 1959, more than 50 years ago. Given the already
high density of the central Bay Area and its environmental constraints, it is clear that land rent
will be a major feature of the housing market for generations to come.