WILL HOUSING BE BUILT?

PGINA 1 DE 11

(http://ternercenter.berkeley.edu)

Housing Development Dashboard

Development Calculator

Top 6 Factors Influencing New Development

We believe that providing a transparent, open resource for policymakers can
improve local decision making. The Housing Development Dashboard shows
how local policies and development factors impact the odds that a housing
development gets built. The Development Calculator focuses on the most important
factors supported by the literature and local development experts. The methodology
and default assumptions were vetted through conversations with area development
experts, data collection, and analysis from January to May of 2016.
The calculator works best for properties of 50 units or more, projects in which the
developer has not yet entered into an option agreement, and where the land seller
is motivated to sell. In reality, many of these factors move together, so users should
be careful to interpret results significantly different from existing market conditions.
The tool is currently in eta testing.

Odds of Construction:

75-100%

25-74%

0-24%

Market Factors
1. Target Return
2. Landowner Willingness to Sell
3. Local Rents and Costs

Local Government Factors

4. Fees or Affordable Housing Requirements
5. Local Planning Decisions
6. Additional Planning Approvals

Other Factors to Consider

Market Factors
Many key inputs to the development process are determined by the market and are fully
or partially out of a local government's control

1. Target Return

HTTP://TERNERCENTER2.BERKELEY.EDU/PROFORMA/

02-08-2016

WILL HOUSING BE BUILT?

PGINA 2 DE 11

Total Project Target Return

11% /year
Odds Of Construction
100%
80%
60%
40%
20%
0%

8%

10%

11%

12%

14%

16%

Most developers do not have the $10s to $100s of millions necessary to fund a typical
real estate project. Outside investors such as a pension fund or Real Estate
Investment Trust provide most of the money for real estate projects and need a certain
return, given the risks involved. Real estate projects are considered risky, and if the
annual rate of return drops to levels close to that of less risky investments such as
Coca-Cola stock or "risk-free" US Treasury bonds investors will choose to put their
money in these less risky alternatives instead of funding this project.
Investors judge risk in a city based on certainty if a number of past projects have been
denied or faced significant delays, investors may require a higher return. Because return
is calculated annually, and compounds over time, target return can make a large
difference.

2. Landowner Willingness to Sell

Market Land Costs

$100 /square foot of land

Odds Of Construction
100%
80%
60%
40%
20%
0%

$0

$100

$200

$300

$400

To calculate whether a project will be profitable, developers often calculate all their costs
the amount they would need to earn to pay back their investors and earn a decent profit,
excluding the price of land. What is left over profits of the project minus costs is
what the developer can afford to pay for land.
Property owners value their property differently, using the value of the existing use the
revenues generated by people paying to park on an empty lot, for example to calculate
how much it might currently be worth. To induce property owners to sell, real estate
developers must often pay more than the property owner's perceived value. Property
owners vary in the value of their current use and financial situation, so in any given area
this value will vary. In a strong market, rising rents will increase the value of the existing
use, meaning a developer must pay more for the land.
That's why in this model, as the market land cost increases, but the amount the
developer is able to pay Land Cost Paid / SF of land, to your right remains the
same, the property is less likely to be built. As the difference between what the developer
is willing to pay and what the landowner sees in the market becomes bigger, the
landowner becomes less likely to sell.

HTTP://TERNERCENTER2.BERKELEY.EDU/PROFORMA/

02-08-2016

WILL HOUSING BE BUILT?

PGINA 3 DE 11

3. Local Rents and Costs

Local Rents

$3.50 /square foot of unit/month

Odds Of Construction
100%
80%
60%
40%
20%
0%

$2.00

$2.50

$3.00

$3.50

$4.00

$4.50

Low Rise Construction Costs

$5.00

$5.50

$6.00

$250 /gross square foot of building

Odds Of Construction
100%
80%
60%
40%
20%
0%

$160

$180

$200

$220

$240 $250 $260

$280

$300

Local rents and sales prices, along with construction costs, make the largest difference in
developer models. Small increases in rents can make a project very profitable, while
small decreases can doom them, and vice versa for construction costs. And as market
rents and sales prices rise, construction costs often rise as well due to skilled labor
shortages, which means that rising rents do not always translate to increased feasibility.
This is one of the reasons development is considered so risky and why investors, and
developers, demand higher returns compared to other investment opportunities. Note this
means that moving one of the levers here may also mean others may shift due to overall
economic conditions, so users should be careful of the conclusions drawn from a large
shift in any one of the indicators.

Local Government Factors

A number of important local government decisions can affect the odds that a project gets
built, especially when these changes are large or the project in question has only a
moderate probability of construction.

4. Fees or Affordable Housing Requirements

Fees

HTTP://TERNERCENTER2.BERKELEY.EDU/PROFORMA/

$30,000 /unit

02-08-2016

WILL HOUSING BE BUILT?

PGINA 4 DE 11

Odds Of Construction
100%
80%
60%
40%
20%
0%
$0

$20K

$30K

$40K

$60K

$80K

$100K

$120K

OR
Affordable Housing

10% of units
Odds Of Construction
100%
80%
60%
40%
20%
0%

0%

10%

20%

30%

Affordability Level

40%

120% of median income

Odds Of Construction
100%
80%
60%
40%
20%
0%

50%

60%

70%

80%

90%

100%

110%

120%

Developers may choose to build affordable housing on site as part of the project or
pay a fee. Compared to local rents and investor risk adjusted return, fees and on-site
affordable housing requirements are not as important to a development model. However,
large changes to the requirements, and the affordability levels required, can have a
significant impact on development in the short term. For example, requiring 20%
affordable housing for lower income people at 50% of AMI or a $40,000 fee makes a
much larger difference than requiring 10% of housing be affordable to higher-income at
120% of AMI or a $10,000 fee.
How a government implements the fees also matters. Developers negotiate an exclusive
purchasing option with a landowner when first putting together plans for a site, which can
happen a year or more before the development begins construction. If a policy does not
grandfather projects during this time period, the developer will have to re-negotiate a
lower price for the land in order to earn a profit. If the landowner refuses to sell at a lower
price, the project may not be built. In addition, investors and developers may percieve the
policy as unfriendly to development and require higher returns in the city in the future.

5. Local Planning Decisions

HTTP://TERNERCENTER2.BERKELEY.EDU/PROFORMA/

02-08-2016

WILL HOUSING BE BUILT?

PGINA 5 DE 11

Height

5 stories
Odds Of Construction
100%
80%
60%
40%
20%
0%

10

20

30

40

Parking

1.00 spaces/unit
Odds Of Construction
100%
80%
60%
40%
20%
0%

0.0

0.5

1.0

1.5

2.0

Basic Permitting Time

6 months
Odds Of Construction
100%
80%
60%
40%
20%
0%

5 6

10

15

20

25

30

Though not in the same realm as local rents and investor risk-adjusted return, local
Planning Department decisions on building height, parking, and permitting time can make
a significant difference to a project's feasilibity. In this case, height refers to the height
chosen by the developer, not a maximum height limit set by the Planning Department, in
order to show the financial tradeoffs at each level of building height instead of just the
most profitable height. Buildings over 6 stories often cost significantly more to construct
because building codes require steel frame instead of the much cheaper wood, and life
safety systems become more complicated. Parking can be very expensive even with
new technology, a single space can cost $25,000 to $75,000. Requiring less parking not
only saves money but can allow for more space to build housing units, and for
developers to earn money.
In the permit process, time is money. Though only a small chunk of the investor's money
is spent up front, the investor still requires an annual return on that sum. The longer the
permit process takes, the more profitable the project must be when it's finally
constructed. Shorter permit times such as shortening building permit wait times or plan
checks can mean significant savings.

HTTP://TERNERCENTER2.BERKELEY.EDU/PROFORMA/

02-08-2016

WILL HOUSING BE BUILT?

PGINA 6 DE 11

6. Additional Planning Approvals

Yes

Additional Approvals (CU or PUD)

No

Requiring that developers obtain Conditional Use (CU) or Planned Unit Development
(PUD) approval can be just as important to developers as local rents or risk adjusted
return. Both are zoning designations created by cities that require a developer to get
approvals from the local Planning Commission and possibly City Council in order to go
through with development. As such, the general public gets to weigh in on the
proceedings, and because the issue is now discretionary, projects must either be certified
as having no environmental impact or go through the process of creating an
Environmental Impact Report.
Issues brought up in these public meetings may be as small as requiring proper signs or
security to appease neighborhood concerns, or they may be as large as providing
additional affordable housing units or mitigating large environmental impacts, such as
traffic issues. In the case of small requests, CUs or PUDs can be relatively harmless, if
they are resolved quickly. In the case of larger issues, developers may end up spending
over $1 million and 12 to 24 months on an environmental impact report, and additional
time in court if sued by neighborhood groups, not to mention the additional costs of
required mitigation.
As a result, investors and developers often require higher returns for properties in these
zones, depending on the perceived risk and additional costs involved.

Other Factors to Consider

Many other factors affect whether a project gets built. For example, different assumptions
for future rent or construction cost increases can radically change a developer's expected
profit. Those willing to take a bigger gamble on larger rent increases could end up big
winners or out of cash.
For Rent

Tenure

Expected Rent Increase

For Sale

3.0% /year
Odds Of Construction
100%
80%
60%
40%
20%
0%

0.0%

2.0%

3.0%

4.0%

6.0%

Expected Construction Cost Increase

HTTP://TERNERCENTER2.BERKELEY.EDU/PROFORMA/

8.0%

10.0%

2.0% /year

02-08-2016

WILL HOUSING BE BUILT?

PGINA 7 DE 11

Odds Of Construction
100%
80%
60%
40%
20%
0%
0.0%

2.0%

4.0%

6.0%

8.0%

Year 1 Return for Similar Properties

10.0%

4.0% annually
Odds Of Construction
100%
80%
60%
40%
20%
0%

2.0%

2.5%

3.0%

3.5%

4.0%

4.5%

5.0%

5.5%

Bank Loan Interest Rate

6.0%

6.0% /year
Odds Of Construction
100%
80%
60%
40%
20%
0%

4.0%

6.0%

8.0%

10.0%

Ground Floor Retail

12.0%

0% of 1st floor
Odds Of Construction
100%
80%
60%
40%
20%
0%

0%

20%

40%

60%

Local Retail Rents

HTTP://TERNERCENTER2.BERKELEY.EDU/PROFORMA/

80%

100%

$25 /square foot of space/year

02-08-2016

WILL HOUSING BE BUILT?

PGINA 8 DE 11

Odds Of Construction
100%
80%
60%
40%
20%
0%
$0

$20 $25

$40

$60

$80

$100

Average Unit Size

$120

800 square feet

Odds Of Construction
100%
80%
60%
40%
20%
0%

400

600

800

1000

1200

Lot Size

25,000 square feet

Odds Of Construction
100%
80%
60%
40%
20%
0%

20K 25K

40K

60K

80K

100K

Liquefaction Zone

Yes

No

Landslide Zone

Yes

No

Sources
This model draws inspiration from the very effective New York Times Buy Rent Calculator
(http://www.nytimes.com/interactive/2014/upshot/buy-rent-calculator.html?_r=0) put
together by Mike Bostock, Shan Carter, and Archie Tse. Key assumptions are drawn from
numerous discussions with developers, contractors, and architects, as well as recent real
estate feasibility studies done in San Francisco, Oakland, and El Cerrito by AECOM
(here (http://files.mtc.ca.gov/pdf/TAP/Oakland_Development_Feasibility_Report.pdf) and
here (http://www.el-cerrito.org/DocumentCenter/View/3699)), Seifel Consulting (here
(http://www.sf-planning.org/ftp/files/plans-andprograms/emerging_issues/tsp/TSF_EconomicFeasibilityStudy_Spring2015.pdf) and here
(http://sf-moh.org/modules/showdocument.aspx?documentid=6976)), and Strategic
Economics (here
(http://www.ci.berkeley.ca.us/uploadedFiles/Planning_and_Development/Level_3_-_DAP/FeasibilityReport_Final.pdf)
Some basic assumptions for affordable condo purchase price calculations are drawn
from the San Francisco Mayor's Office of Housing Inclusionary Housing Program (here

HTTP://TERNERCENTER2.BERKELEY.EDU/PROFORMA/

02-08-2016

WILL HOUSING BE BUILT?

PGINA 9 DE 11

(http://sf-moh.org/modules/showdocument.aspx?documentid=8830)). Structure and basic

assumptions are also drawn from various real estate classes at UC Berkeley taught by
Bill Falik (https://www.law.berkeley.edu/php-programs/faculty/facultyProfile.php?
facID=11321), Dennis Williams (http://facultybio.haas.berkeley.edu/faculty-list/williamsdennis), and Carol Galante (http://ced.berkeley.edu/ced/faculty-staff/carol-galante).

Methodology
The model calculates developer return using the Internal Rate of Return (IRR) metric.
Model assumptions and calculations are based on a pro-forma analysis. For more
detailed methodological information, contact ternerhousing@berkeley.edu
(mailto:ternerhousing@berkeley.edu).
The basic assumptions are as follows:
The condo or apartment project is sold at stabilization after all construction is complete at
a rate of 20 units per month. Stabilization indicates the time when 90% of the building is
occupied for 90 days for rental units.
Hard costs represent costs associated with actually constructing the building digging
the foundation, putting up the walls, installing the plumbing, etc. while soft costs are all
other costs legal fees, some architectural fees, time spent meeting with city staff, and
other costs.
The bank loan is 65% of total hard and soft construction costs and bank loan fees are
0.75% of the loan amount. Costs to build out the retail space are $15 per square foot
(psf) less expensive than low-rise construction costs. Tenant improvements and leasing
costs for the space are $100 psf. Basic podium parking where part or all of the bottom
floor is a concrete-walled parking lot costs $120 psf. Low rise construction cost does
not include retail tenant improvements, parking, or landscaping costs, and indicates the
cost of building up to 5-story structures.
High rise construction costs are calculated internally by the model as a percent of low
rise costs, rising to 103% of low rise costs at 6 stories, 125% at 7 and 8 stories, 130% of
low rise costs at 9 stories, and 1 percentage point higher per story afterward.
Sales expenses are 3.5% of revenue for rental properties and 5.5% of revenue for
condos. There is a premium for condo finishes of 5% of residential hard costs. All units
are the same size in square feet. All units receive the same rent per square foot. There is
a 10% contingency added to hard construction costs.
Each parking space requires 350 square feet, and stackers allowing 2 cars to fit,
stacked one on top of the other, in a single space are used for an extra $15,000 per
space for all parking in buildings 3 stories and higher. With stackers, parking costs come
out to around $28,500 per space in a podium setup, compared to about $42,000 per
space otherwise, a difference of over 30% from traditional parking costs. Construction
takes 18 months for smaller projects 3 stories and higher, 6 months for 1 story projects,
12 months for 2 story projects, and 21 months for projects 7 stories and higher, adding
one additional month time for every 2 additional stories for larger projects beginning at 13
stories.
The amount of lot square footage used as net rentable space is 73% for the most
inefficient projects high rises and near 100% for one to two story buildings.
Landscaping costs $10 psf. Above 50 units, projects gain efficiency at a slow linear rate
up to a 5% discount on hard and soft costs for 250 units and above.
A number of these variables change with building height. Per square foot costs rise with
building height according to general cost estimates from the San Francisco
Transportation Sustainability Fee report and interviews with developers. Lot square
footage used as net rentable space is 100% for 1 story projects and 95% for two story
projects, dropping to 78% for 3 story buildings. The figure improves steadily from 5 to 7
stories, and then declines again after 9 stories in 1% increments. Parking costs per

HTTP://TERNERCENTER2.BERKELEY.EDU/PROFORMA/

02-08-2016

WILL HOUSING BE BUILT?

PGINA 10 DE 11

square feet are significantly cheaper for 1 to 2 story structures than more complex lowrise podium parking. Parking costs rise after 6 stories and again after 11 stories to reflect
the costs of underground parking.
Residential vacancy is 5%, while retail vacancy is 10%. Soft costs, not including interest
carry or local impact fees, are 20% of hard costs. The capitalization rate (cap rate)
termed Year 1 Return for Similar Properties above for retail is 1.5 percentage points
above the residential cap rate, as indicated in a recent CBRE report
(http://www.cbre.us/o/fortlauderdale/AssetLibrary/CapRate_H22014_Master%5B1%
5D.pdf). Two parking spaces are required per 1,000 sf of retail space.
10% of hard and soft construction costs are spent during predevelopment.
Predevelopment indicates the time before construction occurs, usually spent drafting
plans, doing soil studies, getting city approvals, and other activities. 10% of the cost of
the land is paid annually in option consideration.
Basic permitting processes includes time to get a planner assigned, design review, and
time required to receive a building permit application. The EIR or negative declaration
processes are assumed to be concurrent with basic permitting processes, while
conditional use time is assumed to be above and beyond basic permitting time. We
assume developers take about 3 months time before submitting anything to the planning
department.
Projects with a negative declaration are assumed to take 4 months longer than projects
with no environmental review, while projects with a full EIR take 12 to 24 months longer,
depending on project size. EIR production and litigation costs are assumed to start at
$300,000 and rise based on project size by about $1,500 per unit. Projects do not have
to complete an EIR or negative declaration if they are not in a conditional use area, are
not classified as a "large project" here defined using the city of Oakland's definition of
a lot size of 60,000 square feet or more or have 6 units or fewer.
Affordable units are assumed to be the same size and quality as market rate units. AMI
levels are based on 2016 Department of Housing and Urban Development income limits
for a family of 4 in the Oakland-Fremont CA Metro Area. The density bonus is modeled
after Oakland's ordinance 10% minimum required at Low Income, with a 20% bonus,
increasing by 1.5% to a 35% bonus at 20% affordable units. The model uses only the
low-income scale, no matter what affordability level is selected. The model assumes
there is no external subsidy for affordable units, though in practice developers might
pursue a subsidy if more than 20% of units were required to be affordable at 50% of AMI
or below. At this threshold, the project would qualify for non-competitive Federal Low
Income Housing Tax Credit Subsidies.
Projects in liquefaction zones cost 4% more and projects in landslide zones cost 1.5%
more, based on conversations with a structural engineer.
Condo fees are $500/month, and property taxes are 1.2%. Mortgage rates are 1
percentage point below the "Bank Loan Interest Rate" for construction loans. Affordable
condo purchases pay 10% down. Renters and owners pay 30% of their income toward
rent or mortgage payments. Operating costs are 30% of residential rents and 10% of
retail rents.
The land seller in this model is assumed to be a motivated land seller. Land sellers are
assumed to have different prices at which they are willing to sell (for those
mathematically inclined, it is modeled here using a poisson distribution with lambda value
4 at increments of 0.05 from 0 to 10). The distribution, shown below, most closely
matches the distribution of land seller willingness based on conversations with area
developers. This assumes that about 79% of motivated land sellers agree to sell at the
market price, allowing for a few holdouts. The percent of market land value is on the
x-axis and probability of sale is on the y-axis:

Percent of Market Land Value

HTTP://TERNERCENTER2.BERKELEY.EDU/PROFORMA/

02-08-2016

WILL HOUSING BE BUILT?

PGINA 11 DE 11

Odds Of Construction
100%
80%
60%
40%
20%
0%
0%

50%

100%

150%

200%

This model works best at a scale of approximately 50 to 250 units, due to a number of
unique factors in very small and very large projects. At a scale below 50 units, mid-size
projects lose efficiency from a property management perspective, as managers have to
split time between multiple properties. Costs may vary widely for smaller projects as upfront costs easily absorbed into larger projects become a larger share of the project
budget. Single family homes and townhouses at a small scale have different rear yard
requirements, ways of approaching parking, and requirements for building systems.
Project efficiency is assumed to increase by up to 10% of costs as projects rise from 50
to 250 units. Most defaults are set based on local data for a typical mid-sized lot in north
Oakland, though some such as target return and bank loan interest rate are set
based on conversations with developers. Local rents and sales represent the 85th
percentile of north Oakland rents to approximate new construction.
This model does not include an ADU policy as a government lever because the factors
that induce individual owners to build ADUs are too different from the factors facing
larger building developers to include in the model.
2016 Terner Center for Housing Innovation at UC Berkeley. All Rights Reserved. No part of this website or computer code
(javascript, HTML, CSS, etc.) may be reproduced, stored, modified, transmitted, or otherwise used without the prior written permission
of the Terner Center for Housing Innovation at UC Berkeley.

HTTP://TERNERCENTER2.BERKELEY.EDU/PROFORMA/

02-08-2016