B
uilding new housing is complicated and costly. From
land acquisition to project design to entitlement to
nancing, a developer must go through numerous steps
before the rst shovel ever hits the ground. However, the
steps that a developer navigates to take new housing from
idea to occupancy are not well understood by the public or by
policy makers. As a result, some well-intentioned decisions—
such as imposing inclusionary zoning requirements on new
developments when those requirements are not supported
by market fundamentals—may lead to the unintended
consequence of fewer, rather than more, aordable
housing units being built. e “math” underlying housing
development is critical, but very few resources exist to
explain that math to those outside the real estate industry.
We believe that demystifying the math that underpins
whether a project “pencils” is an important step towards
forming a shared understanding of what it will take to move
forward in solving California’s housing crisis. To that end,
the Terner Center has undertaken an analysis of development
scenarios in various regions in Northern California to shed
light on the development process.
is brief will explain the steps a developer undertakes to
design, nance, build, and set the rents for market-rate
housing. We answer the following questions:
» What are the various costs that go into the development
of new housing?
» How are new housing developments nanced?
» What are the benchmarks required by nancial
institutions and capital sources to invest in new housing?
» How do various requirements impact the ability of
developers to deliver projects?
Unpacking the factors that impact what a developer can
build in today’s market helps in understanding why so much
new housing is expensive, and why new supply is oen only
aordable for upper-income households.
1
Terner Center for Housing Innovation • August 2019
1
Making It Pencil: e Math
Behind Housing Development
Copyright 2019 Terner Center for Housing Innovation
For more information on the Terner Center, see our website at
www.ternercenter.berkeley.edu
Authored by David Garcia, Policy Director
Methodology
Underlying every project is a “pro forma”—the analysis a
developer undertakes to estimate total development costs
relative to projected income (e.g., the monthly rents) in order
to determine nancial feasibility. Every type of project—
be it a four-story, 20-unit building or a 20-story, 300-unit
building—will have a dierent cost and return calculus
associated with it. To explain the various elements that go
into a pro forma, we created a prototypical development
project to demonstrate the process by which developers
construct and nance new rental, market-rate housing.
While there are various types of development (e.g., high-
rises, townhomes, accessory dwelling units), our prototype
is designed as a market-rate, mid-rise, rental apartment
building. Specically, the prototype is a multistory residential
building with a Type 1 concrete podium rst oor and
Type 5 wood frame construction above.
2
Since dierent
construction types are subject to very dierent costs and code
requirements, the results of our prototype analysis should not
be extrapolated to other forms of development. For example,
townhomes are oen constructed primarily of wood framing
and do not require more costly features such as concrete
podiums, underground parking, or elevators. Also, high-rise
construction above 85 feet (or roughly seven stories) requires
a shi from wood frame building materials to concrete and
steel, which raises the overall cost of a project considerably.
Regardless of building type, the nancing principles of
any new housing development are the same: any project
must demonstrate the ability to meet an acceptable return
requirement in order to obtain nancing.
To demonstrate geographic dierences in how pro formas
can vary, we assess pro formas for this prototype in three
Total Units
120
Unit Mix
Studio - 48
1 Bed - 40
2 Bed - 32
3 Bed - 0
Parking
120
Ground Floor Retail
1,500 sf
Terner Center for Housing Innovation • August 2019
2
dierent areas of California: the East Bay (e.g., Oakland,
Berkeley), the South Bay (e.g., San Jose, Santa Clara), and
Sacramento. As with design type, costs can vary signicantly
by region which may feature dierent building codes, local
requirements, and labor markets. e selection of these
regions was informed by our project partners who were
interested in the costs to develop in their specic localities.
3
To determine project costs, we put the hypothetical prototype
“out to bid” with general contractors to provide a broad
estimate of how much this project would cost in today’s
market. Our analysis was also informed by conversations with
developers in each region to ensure that project assumptions
and characteristics were representative of each market.
To determine whether the prototype would “pencil” in
each market area, we also made a series of assumptions
regarding the nancial thresholds each project must meet.
As with project cost, these assumptions (explained in more
detail below) were vetted with lenders and equity providers
and represent standards in their respective market areas.
In addition to the overall cost of construction, nancing
dynamics drive rental costs, as the obtainable rents for a
project must support the overall cost to develop, while also
meeting the nancial requirements of banks and investors.
Beyond our baseline assumptions, we also examined how
changes in pro forma inputs, such as higher impact fees
or reduced parking requirements, changed our project’s
calculus and potential for feasibility. is additional
analysis was done to explore how dierent policy priorities
are reected in the development math.
Table 1: Prototype Characteristics
Terner Center for Housing Innovation • August 2019
3
No Environmental Impact Report
e prototype is not required to conduct a full Environmental
Impact Report, as generally required by the California
Environmental Quality Act (CEQA).
No Aordable Housing Requirement
e prototype is not required to include any below-market
units, or pay into an aordable housing program.
No Demolition
No existing structures existed on the site that required
demolition.
Total Impact Fees of $40,000/unit
Impact fees are fees levied on a project as a condition of
approval by a city, county, or other fee-levying body (e.g.
school district, municipal utility district).
No Environmental Remediation
e prototype site does not require any signicant
remediation of contaminated soil, or other issues commonly
found in urban inll locations.
Parking Requirements
One parking space per unit is required.
No Signicant Osite Requirements
ere is sucient existing infrastructure to service the
prototype. e project does not need to undergo signicant
work in order to improve capacity for services such as water,
power, or wastewater.
Non-Prevailing Wages
Prevailing wages for the prototype are not required as no
subsidy is used for project nance.
5
Table 2: Prototype Assumptions
Project Prototype
While in reality no two housing developments are the
same,
4
we developed a simple prototype project in order to
show how the math works and how the same project will
have dierent nancial fundamentals in dierent housing
markets. Table 1 depicts the design characteristics for our
prototypical project including unit number, mix, parking,
and retail.
In addition to specic development characteristics, we also
made a series of assumptions regarding site conditions as
well as jurisdiction requirements (e.g., parking require-
ments). ese assumptions are detailed in Table 2.
While we made these assumptions in order to compare
across prototypes, it should be noted that any increase
or change to any combination of these components
could dramatically increase the total cost of a project.
For example, land that requires signicant remediation
of contaminated soil or the demolition of an existing
building would increase the total cost of development. City
requirements could signicantly change the development
math as well, such as requirements to upsize underground
infrastructure, or provide signicant o-site, public right-
of-way improvements as a condition of approval.
5
Breaking Down Costs
ree categories of costs are associated with any development
project: hard costs, so costs, and land costs. We’ve broken
down our project prototype by these three categories.
Hard Costs
e most signicant costs of any project are those associated
with its physical construction. ese include labor and
materials, including the cost of concrete, timber, and
Terner Center for Housing Innovation • August 2019
4
mechanical systems. Hard costs are also reective of various
building code requirements that impact the way a structure
is built. For example, in seismic zones, building codes
require new buildings to be constructed with materials that
will hold up during an earthquake. A 5 percent hard cost
contingency is also included in our pro formas to mitigate
against project overruns.
So Costs
e second largest component of overall project costs are
known as “so costs,” which are those associated with
the design and implementation of the project, but not the
physical construction (i.e. hard costs). As there are numerous
components to any project’s so costs, we’ve grouped them
together in broad categories—fees, nancing, consulting,
and tax, title, and insurance—each of which is described in
more detail in Table 3. We assumed a at rate of $40,000 in
impact fees per unit for the purpose of comparing equally
across markets. However, it should be noted that fees vary
widely by jurisdiction, with many localities charging much
more than our assumed amount.
6
Consultant costs refer to the broad set of experts that
developers need to design and execute a building. e
majority of these costs go to the architects and engineers
hired to work on the project. We also include the costs
of the various other consultants typically required for
a development of this size. On any given project, these
consultants can include, but are not limited to, geotechnical
engineers (e.g., to determine the integrity of soil on the
project site); historical resource consultants (e.g., to
determine whether the site contains buildings or resources
that can be considered historic); and joint trench consultants
(e.g., consultants that coordinate trenching of existing
overhead utilities). It should also be noted that the bulk of
these costs are typically paid for upfront by the developer
before obtaining nancing or full project approvals.
e nancing category includes items that are related to the
costs associated with obtaining nancing, but do not include
return requirements, which are discussed in the next section.
e nancing of so costs includes a required interest
reserve (to pay interest on the construction loan during the
construction period), so cost contingency funds (in case
so costs exceed the budgeted amount), and the costs to close
nancing. In addition, the nancing section also includes an
amount for a “developer fee.” Generally, nancial partners
allow developers to budget 3 to 5 percent of the total project
cost in order to compensate for risk and cover developer
overhead, which includes the out-of-pocket expenses that
the developer incurs specic to managing the project during
predevelopment, construction, and lease-up phases.
7
is
fee is not accessible to a developer until nancing has closed
and construction commences. As such, costs associated with
paying development team sta, city fees for plan check and
other services, and general property holding costs are borne
solely by the developer.
Fees
Includes any fees required as a condition of approval for the
project, such as school fees, utility connection fees, park fees,
art fees, or transportation fees. NOTE: our prototypes stan-
dardized total fees at $40k per unit, however total fees vary
widely by jurisdiction, and are levied by dierent entities.
Consultants
Includes costs associated with professional services to design
the project. is includes, but is not limited to, architects;
structural engineers; civil engineers, landscape architects;
mechanical, electrical, and plumbing design; geotechnical
engineers; joint trench consultants; waterproong consul-
tant; accounting; and legal.
Financing Costs
Includes costs associated with obtaining debt and equity,
including loan closing costs, so cost contingencies, and
operating reserves. Also, our prototypes include 3 percent
of total costs for a “developer fee” to mitigate developer risk
and pay for overhead to build and manage the project.
Tax, Title, and Insurance
Includes costs of general liability and builder’s risk insur-
ance, as well as property taxes during construction.
Table 3: So Cost Categories
Terner Center for Housing Innovation • August 2019
5
Land Costs
e cost of land is a signicant part of overall costs, but is
determined in a dierent manner from hard and so costs.
Generally speaking, land costs are “residual” in that the cost
of land should be determined by the amount a developer
can aord to pay for the land without making the project
too expensive to reach threshold nancing requirements.
Put another way, the cost of land should be determined by
the amount of funds le over aer estimating total hard and
so costs without pushing the project into infeasibility. In
theory, the market value of land, and what the developer can
pay for it, is driven by what can be developed there.
In reality, however, land costs are impacted by various
factors, many of which are not related to project feasibility.
For example, a property owner may hold out on selling
property at the residual price to a developer for many
reasons, such as: continuing to operate a protable business
on the property (e.g., a surface parking lot), anticipating that
the value of the land will increase in the future, or owing
more on the property than the residual value. In these and
other instances, a residual land price oered by a developer
may be less than what a property owner is willing to sell for.
As a result, developers must choose to pay more than the
residual value or not purchase the land at all.
For our prototype, we determined land cost by using
comparable sales of land in each of the three markets.
In addition to total cost, this category also includes costs
associated with closing on the land, as well as due diligence
reports (e.g., environmental “phase 1” or “phase 2” reporting
to determine the extent to which the presence of harmful
substances exist on the site).
Terner Center for Housing Innovation • August 2019
6
Pro Forma Cost Results
Based on the characteristics and assumptions described above, we calculated the total cost of the prototype
in the three markets:
• Costs were highest in the East Bay, with the project estimated at a total of $68,828,255 ($573,569/unit),
driven by higher construction and land prices.
• e South Bay project was similarly costly at $61,579,785 ($513,165/unit).
• In Sacramento, the prototype was signicantly less expensive at $45,581,075 ($379,842/unit) due to
lower construction and land costs.
Figure 1: Total Prototype Cost
Figure 2: Total Per Unit Prototype Cost
Terner Center for Housing Innovation • August 2019
7
Market Study
Developers must provide a study by a reputable consulting
rm demonstrating sucient demand for housing at the
project’s projected rental prices.
Environmental Documents
Lenders will request documentation identifying any existing
environmental issues with the site, such as the potential for
contaminated soil from an underground gas tank or well.
Appraisal
e appraisal determines the project’s market value upon
completion.
Architectural Documents
Lenders require architectural plans from the developer,
including engineering and other technical documentation
necessary to construct the building.
Approvals
In order to close a construction loan, the project must have
obtained all approvals from relevant governing bodies. is
includes entitlements and building department approvals
from the local municipality, but also any approvals required
from other agencies with jurisdiction over development in
the project’s location (e.g. regional water quality control
board, or air resources board).
Table 4: Examples of Lender Documentation Requirements
Project Financing
To pay for the cost of these prototypes, a developer will obtain
funding from two sources: debt and equity. Debt provides the
bulk of project nancing, while equity provides the balance.
Both forms of funding have their own strict thresholds
and requirements that a developer must meet in order to
obtain money to build the project. ese requirements also
inuence project feasibility, and can add to the overall costs
of development.
Debt
Debt is provided in the form of a loan from a lender (generally
a bank) and carries an interest rate which the developer pays
back over time.
8
Interest rates vary across market cycles, but
for the purposes of this brief, we’ve assumed a total interest
rate of 5 percent across each prototype.
Lenders examine two components when considering whether
or not to provide a loan to a particular development: the
developer capacity and the details of the project.
9
Developers
must show that they have a proven record of success in
completing projects on time and on budget, and paying back
debt. Lenders also oen require a developer to personally
guarantee the project loan, which puts tremendous risk
on a developer, and severely limits developers who do not
have the personal assets to sign such a guarantee. In other
words, if the project doesn’t succeed, the developer is oen
personally liable to repay the lender.
Lenders also require supporting documentation to ensure
that the project will be successful, and will not agree to
fund a project or release funding until this documentation
is provided. is includes but is not limited to: market
studies, appraisals, environmental documents, architectural
documents, and approvals from localities and agencies that
have jurisdiction over development in the project’s location
(Table 4). Developers must provide these at their own
expense and risk before closing on project nancing.
10
In addition to assessing a developer’s track record and
a project’s documentation, banks require a project to
demonstrate the ability to meet certain nancial benchmarks.
ese nancial benchmarks help a bank to determine the
likelihood a project will not go into default—that projected
long-term income on the project will more than cover the
payments on the loans that the bank would make on the
development. While there are many benchmarks required by
dierent institutions, we focus on two specic metrics: debt
service coverage ratio (DSCR) and loan-to-cost ratio (LTC).
Terner Center for Housing Innovation • August 2019
8
Debt Service Coverage Ratio (DSCR)
To mitigate risk, a bank requires a project to demonstrate
that its income can support the monthly loan payments over
the life of the loan. is metric is measured by a debt service
coverage ratio (DSCR) and is calculated by dividing the
project’s net operating income (NOI)
11
by the anticipated
loan payment. For example, a projected DSCR of 1.0
indicates that a project anticipates achieving exactly enough
income to match what is required to pay its debt. However,
banks require the DSCR to be higher than 1.0 for real estate
lending to ensure that, if NOI projections are inaccurate, the
developer can still meet their debt obligations. For instance,
in the case of a DSCR of 1.0, any small change in NOI–such
as higher than anticipated maintenance costs, or lower rent
revenues–would put the developer in danger of not meeting
their debt obligations. We have made the assumption that a
lender would require a DSCR of 1.3, meaning that our projects
must demonstrate an NOI 1.3 times the amount of debt
issued. While this ratio is standard in California, it could be
more or less depending on the specic lender’s requirement.
Loan to Cost (LTC)
While the majority of funding for a standard market-rate
project takes the form of debt, banks do not provide loans
on the total cost of a project. To further minimize risk, banks
require developers to bring in equity for the amount of the
project that is not covered by the loan (described in detail in
the following section). is is similar to a traditional home
mortgage where a bank requires the buyer to make a down
payment of 20 percent of the value of the house.
e amount that banks are willing to lend relative to the total
project cost is referred to as the loan-to-cost ratio (LTC)
(Figure 3).
12
Lower LTC ratios indicate lower condence that
a project will perform as anticipated given market conditions
and trends, while higher LTC ratios indicate stronger
condence in project success. Typically, in California,
lenders currently require an approximately 65 percent LTC
ratio. We use this ratio in our analysis, but as with the DSCR,
the LTC ratio can vary by region, project, or bank.
Equity
Aer determining how much debt can be obtained, each
prototype is le with a “gap” between the total cost of devel-
opment and how much of the project can be nanced with
loans. is gap is lled by equity, which comes from a project
investor (as well as a smaller amount of equity provided by
the developer). It is important to note that this form of capital
is not specic to real estate, and can ow to wherever it can
achieve the highest risk-adjusted rate of return.
Equity investors consider housing development a riskier
investment when compared to traditional forms of
investment such as stocks or bonds. Because of the length
of time needed to develop before revenue is generated, as
well as the myriad challenges that new housing faces (e.g.,
unpredictable costs, market cycles, construction delays),
there are many variables that could lead a project to deliver
below expectations or be derailed entirely. Investors in real
estate account for these risks by requiring certain levels of
Figure 3: Amount of Debt Compared to Cost for ree Regions
Terner Center for Housing Innovation • August 2019
9
return, and as a result developers must demonstrate that
they can achieve sucient returns to justify the risk.
Equity investors in residential real estate come in various
forms, and are not exclusively private equity groups.
Depending on the size and experience of a developer, private
equity is obtained from insurance companies, foreign capital,
and the pension funds of public sector employees and union
members that utilize real estate development investments as
“high return” options to round out their overall portfolios.
is means developers are beholden to equity returns in
the market, which are set as much by Wall Street as by local
conditions, and oen do not relate to how much prot a
developer makes from a project.
A developer gains equity by contributing a portion of their
own capital (in the case of larger developers), as well as
through their time to develop and manage the project (also
known as “sweat equity”). e percentage of developer
equity is generally a much smaller percentage than that of
the investors. Prots received by a developer are not realized
until at minimum the loan payment has been made, and the
investor has received their preferred return. In most cases,
developers will not see prot until equity is fully returned
to investors. is is a key point to understand as a developer
is generally the last stakeholder to receive any prot from a
new housing development, and most developers shoulder a
signicant amount of risk and cost even before any form of
nancing is secured.
Measuring Return
e form of equity nancing is critical in determining
whether a project gets built. Ultimately, a developer will
make a decision to build or not based on whether they can
achieve threshold return requirements that will allow them
to attract equity. While there are several ways to measure
return, the simplest metric is to compare a project’s antici-
pated return-on-cost (ROC) to local area capitalization rates.
e ROC can be used to compare returns across various
investment types. With regards to real estate development,
ROC measures the expected return aer accounting for
the cost to build and manage a new housing development.
is metric is determined by dividing a project’s anticipated
NOI by total project cost. Capitalization rates, on the other
hand, measure the return one can expect by purchasing a
certain property. Essentially, by comparing ROC to capital-
ization rates, a developer is measuring the return of building
a new project against the return of simply buying an existing
building. If the project’s ROC is reasonably above the capi-
talization rate, then a developer will move forward. To put
it another way, a developer will not go through the time
and expense of developing a new project if it will not yield a
higher return than they would receive by buying an existing
property in the area.
We use this ROC to capitalization rate comparison to deter-
mine feasibility for each of our projects. e extent to which
a project ROC must surpass capitalization rates to achieve
feasibility changes according to the region, project type,
and investor (including their views on timing relative to the
market cycle). To determine this variable, we spoke to devel-
opers, consultants, and architects in each region. Based on
these conversations, we determined that a minimum spread
of between 1.0 percent and 1.5 percent is needed for projects
in the East and South Bay regions, while projects in Sacra-
mento are moving forward at a spread of 1.5 to 2.0 percent.
ROC is determined by dividing a project’s Year 1 NOI by total
project cost. As illustrated in Figure 4, each project’s ROC
varied to a degree. However, these project ROCs all reach
our threshold requirements for feasibility when compared
to area capitalization rates.
13
For our East Bay project, the
ROC is 1.16 percent percentage points higher than area
capitalization rates for new buildings.
14
Similarly, our South
Bay project achieved a spread of 1.14 percent between the
project ROC and capitalization rates. For Sacramento, our
project demonstrates a 1.51 percent spread between ROC
and capitalization rates. While these three projects each
reached the minimum threshold requirements as explained
above, they fall on the low end for feasibility, with little to no
room for additional cost increases.
Beyond ROC, investors use other metrics to determine
their interest in a project. For example, another common
metric is internal rate of return (IRR). IRR measures an
investor’s total anticipated return over the life of their
investment (as opposed to the Year 1 return, as measured
by ROC). Specically, the IRR is calculated by summing
the anticipated annual cash ow for the number of years an
investor expects to hold the property (generally 10 years)
with the anticipated value at sale. Depending on the type of
investor, IRR requirements can uctuate signicantly. For
example, some investors will only invest in projects whose
IRRs exceed 20 percent (e.g., a high-yield investment fund)
while other funds may be comfortable with projects with
IRRs closer to 15 percent. e IRRs demonstrated by our
projects are 15.4 percent for the East Bay, 15.2 percent for
Terner Center for Housing Innovation • August 2019
10
Figure 4: Prototype ROC/Cap Rate Spread
Figure 5: Prototype Internal Rate of Return
Terner Center for Housing Innovation • August 2019
11
the South Bay, and 18.2 percent for Sacramento (Figure 5).
ese IRRs would be attractive to some investors, but not all,
which limits the available pool of capital for the developer.
Project Rents
New developments derive the vast majority of their income
from rents charged to tenants.
15
To determine rents in new
projects, developers must commission a detailed market
analysis from private consultants. ese consultants use
proprietary data sources to determine the demand for new
housing in the project area as well as what a developer could
expect to receive in rents. A developer uses these numbers
to complete their pro forma and to prove to lenders and
investors that their project will receive enough income from
rents to justify their nancing of the project. If developers
cannot produce evidence that they can achieve rents high
enough to satisfy both lender and investor requirements,
they will not receive nancing.
Rents for each of our projects were determined by assessing
area rents in similar new projects. ese rents are illustrated
Figure 6: Prototype Monthly Rents
below in Figure 6. In each case, the rents required to make
the project feasible are higher than what most renter house-
holds in each region can aord. When compared to income
levels of renter households in each region, the minimum
rents required for a two bedroom unit are only aordable to
those with the region’s highest incomes.
16
Layering Requirements
What happens when projects face additional costs, either
due to unknown cost factors (e.g., learning that the land
has soil contamination that requires remediation) or local
requirements, such as parking, inclusionary zoning, and
development impact fees? Anything that drives up project
costs will aect the pro forma calculations and inuence
whether the project is nancially feasible.
To illustrate this dynamic, we have layered a handful of
common requirements onto each of the three project pro
formas. Specically, we added three inputs: increased parking
(2:1 ratio of parking spaces to units); higher fees (total of
$60,000 per unit); and a 15 percent inclusionary zoning
Terner Center for Housing Innovation • August 2019
12
requirement (at 60 percent of AMI). ese variables were
selected because they can have a signicant impact on overall
cost, and also vary widely across—and even within—cities.
Cities have sole authority to determine these requirements
(with the exception of development fees levied by other
entities, such as school districts and utility districts). We
compare the eect of these changes against the capitalization
rate; as explained above, projects that fall close to or below an
area’s capitalization rate are unlikely to be built.
As demonstrated in Figure 7, the combination of these three
requirements brings each project well below the minimum
threshold for viability. In these cases, rents would need to be
signicantly higher in order to maintain viability. However,
given that rents for the baseline prototypes are already only
aordable to those with the region’s highest incomes, it is
likely that the market would not be able to support the rents
necessary to absorb these additional requirements, and the
project would not be built.
On the other hand, reducing costs in other areas can
allow projects to “pencil” while achieving policy priorities.
In other words, a priority like inclusionary zoning may
become viable if, for example, a jurisdiction reduces parking
requirements. To demonstrate this concept, we adjusted a
handful of cost assumptions by: reducing total fees ($20,000
per unit); reducing parking requirements (0.5:1 parking
spaces to units); and tax exempting aordable units. As
illustrated in Figure 8, each project falls below the return on
cost threshold for feasibility when a 15 percent inclusionary
zoning requirement is layered onto each project’s baseline
pro forma. However, as osets that help reduce overall costs
are added, each project moves back towards feasibility to the
point where project returns are actually healthier than the
original pro formas without inclusionary units. By calibrating
policy to account for the overall cost to build, policy makers
can enable developers to build projects that meet important
policy goals, such as including on-site aordability, without
jeopardizing their ability to obtain nancing. e options
presented here are not the only opportunities to provide
meaningful osets as policies such as approval streamlining,
which limits cost increases and holding costs (e.g.,
maintenance, property taxes, insurance), additional density
or height bonuses that allow for more units, and exibility
on aordability (i.e. allowing higher AMIs for inclusionary
units) can also impact project feasibility.
Conclusion
Building new housing is complex and costly, and
understanding the process developers follow to build
housing is important for determining appropriate policy
responses. As demonstrated by our pro forma analysis, there
are dozens of inputs and requirements that directly impact
the cost to build new housing and the amount at which new
housing can be oered to renters. Even where development
conditions are favorable, the overall cost to build makes it
dicult for developers to deliver housing at price points
aordable to lower-, middle-, and increasingly upper-
middle-class households in high cost regions in California.
Given the complexities and costs involved with creating new
housing, policy makers at all levels of government should
be cognizant of how requirements interact with the math
behind housing development. While many requirements
are intended to help achieve important policy objectives—
creating deed-restricted aordable housing or expanding
park facilities, for example—they may inadvertently
push new housing into the red. oughtful approaches to
balancing various priorities are required to ensure that these
policies can work with new housing development rather
than against.
e Terner Center has shown that broad tools can be
created to provide these important insights. In addition to
the pro forma analyses presented in this brief, the Housing
Development Dashboard allows users to design a “prototype”
project using various market and policy inputs. Using these
inputs, the dashboard then determines the likelihood that
the project will be built, using similar nancial assumptions
as presented in this brief.
As local, regional, and state policy makers consider various
policies for increasing housing supply and aordability, tools
such as the Terner Center Dashboard or other thoughtful
and technical approaches should be utilized and expanded
to provide insight into the relationship between new housing
development and other important factors. Embedding this
type of analysis is integral to understanding the potential
outcomes of various policy choices.
Terner Center for Housing Innovation • August 2019
13
Figure 7: Cumulative Impacts of Additional Requirements on Prototype Pro Formas
Terner Center for Housing Innovation • August 2019
14
Figure 8: Cumulative Impacts of Inclusionary Zoning Osets on Prototype Pro Formas
Terner Center for Housing Innovation • August 2019
15
1. While this brief does not delve into the factors driving increasing construction costs, future Terner
Center research will address this specic issue.
2. is type of construction is commonly referred to as “ve-over-one”.
3. e selection of the East Bay was driven by the Center’s work with the Metropolitan Transportation
Commission, specically on the Committee to House the Bay Area (CASA) Initiative; the selection
of the South Bay was driven by support from SV@Home; the selection of Sacramento was driven by
support from the Sacramento Area Council of Governments.
4. Even within a similar building type, development characteristics are dictated by lot size and shape,
as well as local city zoning standards, such as setbacks, lot coverage, and allowable density or height,
among others.
5. Prevailing wage requirements mandate that any contractor working on a project pay a predeter-
mined wage rate. ese wages are generally required on projects that utilize public subsidies, such
as with many aordable housing projects, or in instances where the developer has negotiated with
local stakeholders to pay prevailing wages.
6. Mawhorter, S., Garcia, D., & Raetz, H. (2018). It All Adds Up: e Cost Of Housing Development
Fees In Seven California Cities. Berkeley, CA: Terner Center for Housing Innovation. Retrieved
from https://ternercenter.berkeley.edu/development-fees
7. Peiser, R. B. & Hamilton, D. (2012). Professional Real Estate Development: the ULI Guide to the
Business. ird Edition. p. 213-214. Urban Land Institute
8. ere are two types of loans necessary for rental housing projects: a Construction loan, which is
the loan used to pay for the construction of the project, and Permanent debt, which is the long-
term mortgage on a nished project. Once construction is complete, a lender will “take out” the
construction loan (i.e. pay o) and replace it with a xed-rate permanent loan. e permanent debt
cannot be obtained until the project has been completed and reached nancially sustainable occu-
pancy at rent levels that support the debt.
9. Peiser, R. B. & Hamilton, D. (2012). Professional Real Estate Development: the ULI Guide to the
Business. ird Edition. p. 64-65, 209. Urban Land Institute
10. e loan will also include a “not to exceed” amount, as well as contingency amounts for unforeseen
issues (i.e. cost overruns).
11. Net Operating Income is dened as project income derived from rents minus expenses of operating
the property (i.e. maintenance, leasing, property taxes, legal, sta) before paying debt.
12. Loan to Value is another metric by which banks will measure the amount they are willing to lend.
While similar to LTC, LTV sizes the loan relative to the nished value of the project, rather than the
project cost.
Endnotes
Terner Center for Housing Innovation • August 2019
16
13. Capitalization rates for the East Bay and South Bay regions were determined through an analysis of
Yardi data. Capitalization rates for Sacramento were pulled from CBRE’s North America Cap Rate
Survey, Second Half 2018.
14. ese percentages are commonly referred to as basis points, or BPS.
15. Income is also derived from parking charges, retail rent, and other sources, if applicable.
16. “Aordable” is determined by calculating 30 percent of income for housing costs.
Terner Center for Housing Innovation • August 2019
17
Acknowledgements
We would like to thank the Metropolitan Transportation Commission, SV@Home, and the
Sacramento Area Council of Governments for the funding and support that made this work
possible. e author would like to thank Charmaine Curtis and Dennis Williams for their
insights, as well as Carol Galante, Carolina Reid, and Elizabeth Kneebone for their review.
