CHAPTER 1 BRIDGING THE GOLDEN GATE:

OUTSOURCING TO A NEW PUBLIC ENTITY

Infrastructure Development Systems IDS-OO- T-O 16

Research Assistants Bradley Moriarty and Kai Wang prepared this case under
the supervision ofProfessor John B. Miller as the basis for class discussion, and not
to illustrate either effective or ineffective handling of infrastructure development
related issues. Data presented in the case has been altered to simplify, focus, and to
preserve individual confidentiality. The assistance of the Golden Gate Bridge,
Highway and Transportation District is gratefully acknowledged. To learn more
about the project, the District's publication "The Golden Gate Bridge: Report of
the Chief Engineer" is an invaluable resource.

Bridging the Gap

San Francisco thrived throughout the 1800's as a popular and busy

port. With the arrival of the trans-continental railroad in Oakland in the
1860's, San Franciscans watched port related commerce spread out of their
city toward Oakland. The only rail access to San Francisco was by the
Dumbarton crossing far to the south. Although San Francisco's importance
as a financial center continued, talk about establishing a fixed crossing of
the bay to San Francisco led nowhere.

In the 1900's, the automobile created the rise of fixed, different and
greater, pressures for a bridge. By 1918, city planners were considering
spans across San Francisco bay as an alternative to the existing ferry system.
A system of antiquated ferries was operated by the Southern Pacific
Railroad without competition and was little more than a collection of boats
jerry-rigged to carry automobiles. The vessels crossed the bay slowly,
infrequently, and charged high tolls.

Beginning in 1921, several independent ferry companies took

advantage of high demand for travel across the bay by providing fast and

J. B. Miller, Case Studies in Infrastructure Delivery

© The Massachusetts Institute of Technology 2002
2 Case Studies in Infrastructure Delivery

frequent service on diesel powered ships specifically designed to carry cars.

This new service, with reduced tolls, quickly gained market share, and
customer demand steadily increased along with new capacity. Attempts by
the Southern Pacific Railroad to buyout these new operators created further
interest in automobile travel across the bay. In May 1929, Southern Pacific
briefly regained control of all cross-bay ferry service by purchasing all of
the operators and merging them together. Travel on the ferry system had
increased by over 700% in one year.

The unification of the ferry system allowed Southern Pacific to better

match capacity to demand, further improving overall system capacity. Even
so, by the time of the merger, demand had again expanded beyond the new
capacity. Sunday night traffic jams on the highways of Marin County
leading to the Sausalito ferry terminal attested to the pent-up demand for
travel across the Golden Gate. As Southern Pacific continued its efforts to
control and expand the ferry business, the municipal government of San
Francisco passed local ordinances that would make a bridge across the
mouth of the San Francisco Bay - the Golden Gate - possible.

Location of the Bridge

Geographically, the mountains of Northern California divide the state

into long north-south valleys and intermittent high and low passes. The
action of the rivers, now called the Sacramento and the San Joaquin, carved
out the pass that now forms the Golden Gate. This allowing the Pacific
Ocean to flood the 463 square miles of basin, leaving the peninsula of
present day San Francisco surrounded on three sides by water (See Exhibit
I-Ion the companion CD). The valley that ran north-south though Marin
County to San Francisco provided a natural crossing at the Golden Gate. A
bridge crossing at such a location forms a natural geographic monopoly to
access San Francisco from the north.

Exhibit 1-1 Map o/San Francisco Bay Area

Exhibit 1-1 By Permission, Golden Gate Bridge and Highway District, from The
Golden Gate Bridge, Report of the Chief Engineer, 50th Anniversary Edition, 1987.
 Chapter 1 Golden Gate Bridge 3

Assessing the Bridge's Feasibility

Studies for the Golden Gate Bridge began with a 1918 City Council
resolution ordering the feasibility study of a bridge across the Golden Gate.
The city engineer, M. M. O'Shaughnessy, approached Joseph B. Strauss, a
nationally recognized bridge engineer to look at the project. O'Shaughnessy
was not convinced that the project was technically or economically viable.

"Everybody says it can't he done and that it would cost over

$100,000,000.00 if it could he done."!

Strauss disagreed and thought that a suitable bridge could be

constructed for a more affordable price after carefully studying site
conditions at the Golden Gate.

Following a study jointly undertaken by the City and the United States
Coast & Geodetic Survey,2 a Request for Proposals (RFP) was sent out in
May 1920 from O'Shaughnessy's office soliciting designers and cost
estimates for the project. The RFP was sent to Joseph Strauss in Chicago,
Francis C. McMath of the Canadian Bridge and Iron Company of Detroit,
and Gustav Lindenthal, the engineer for the WOO-foot Hell's Gate Bridge
over the East River in New York in 1916.3

McMath never officially responded. Lindenthal estimated that the

bridge would cost between $60 and $77 million dollars. Strauss, however,
estimated a total cost of $27 million. Strauss' estimate, although over the
$25 million budget proposed by the city, brought the discussion from
theoretical possibility to serious consideration. 4 Strauss was selected as the
Project Engineer.

Among the many obstacles facing the bridge at that point, the most
important were the rail/ferry monopoly and the military, which owned the
land at each end of the proposed bridge. Southern Pacific was then in the
midst of its effort to regain its share of the ferry traffic from the independent
operators. Southern Pacific was investing heavily in improved ferry
services and was loath to let a fixed link siphon off all ferry customers.
Exhibit 1-1, on the companion CD, shows the results of a 1935 traffic study
which projected traffic flows that might use the new bridge.
4 Case Studies in Infrastructure Delivery

Exhibit 1-2 1935 Traffic Flow Diagram

Exhibit 1-2 By Permission, Golden Gate Bridge and Highway District, from The
Golden Gate Bridge, Report of the Chief Engineer, 50th Anniversary Edition, 1987.

The Army and Navy had different concerns. First, the approaches to
the bridge would, by design, pass through the Army's Presidio base. The
Navy wanted to insure that its major naval base in San Francisco would
continue to have access to the Pacific Ocean.

Because the use of military land was crucial to the success of the
project, one of the first steps taken in the planning process was to apply to
the Secretary of War for a pennit to build the bridge and make use of federal
land needed for the bridge structure and approach roads. The application
was filed by the "Bridging the Golden Gate Association" and
representatives from San Francisco and Marin Counties before the group
had official authority to build the bridge. Colonel Herbert Deakyne held a
hearing on the petition in May 1924.5 Based on Deakyne's favorable report,
the Secretary of War, Weeks, issued a provisional pennit in December 1924
to proceed with the bridge, pending submission of detailed plans. 6

The Formation of the Golden Gate Bridge Highway and

Transportation Authority
The Bridge and Highway District Act created the Golden Gate Bridge
Highway and Transportation Authority (the "Authority"), authorized
multiple counties on both sides of the Golden Gate to vote to allow the
District to borrow money secured by taxes on private property in the
supporting counties. The ability to tax would defray the initial expenses and
provide lending institutions with valuable assets to secure debt repayment if
the toll revenue failed to meet the bond payment obligations. The Act was
written so that citizens of the counties within the District could, by vote of
each county, include or exclude each county in the District. (See Exhibits 1-
4 through 1-6 on the companion CD for infonnation on the bridge bonds)

Exhibit 1-3 Summary ofthe Bridge and Highway District Act of California
 Chapter 1 Golden Gate Bridge 5

After the 1924 Act was passed, forces in opposition to the bridge
brought suit in every county listed in the District: San Francisco, Marin,
Sonoma, Napa, Humboldt, Mendocino, and Del Norte. The litigation
continued for nearly six years, during which time Humboldt County and
portions of Mendocino County dropped out as members of the District. The
factions opposed to the bridge argued that it would be far more expensive
than Strauss' $27 million estimate and that the tolls would never be
sufficient to pay for interest and principal on the bonds sold to finance the
initial construction and for maintenance of the structure. The claim was that
taxpayers in the District would have to pick up the difference; i.e. that
individual property owner in the various counties would be taxed to retire
the debt. Another argument against the bridge was that it would slow future
development of San Francisco by limiting the height of ships entering the
harbor to 200-feet.

On December 4, 1928, after the first round of litigation was concluded

in favor of the District, the District was formally incorporated under the
1924 Act and the process of getting approval from member counties to issue
the bonds needed to fund the bridge began.

Bidding the Project

The District's first action, in July 1929, was to increase the tax rate of
3¢ for every $100 in taxable property in the District. The purpose of the tax
was to fund planning and early design of the project. The tax fueled new
opposition arguments. "Taxpayers forced to outrageous payments for the
bridge." The arguments were forcefully made by experts and men of
standing in the community. The District responded by pledging that the
entire project would not cost more than $35,000,000. Just months later, in
November 1930, votes in the District approved the first bond issue to
finance the bridge by a vote of 145,057 to 46,954. 7

To increase the level of confidence that the entire project could be built
in the proposed budget, Strauss obtained firm fixed quotes for all major
components of the project. This practice was unusual in 1930, but
commonplace today.
6 Case Studies in Infrastructure Delivery

"Pursuant to their pledge, therefore, the Directors ordered the Chief

Engineer (Mr. Strauss) to prepare plans and specifications covering all units
of the work so that lump sum bids could be received and the construction
cost of the project ascertained in advance of the awarding of any contracts.
Such procedure, of course, is contrary to standard practice on major jobs,
where the starting of certain units of construction must, in the nature of
things, be postponed until other units have been finished."8

Strauss included an early version of the "Changes" clause that is now

standard in public construction contracts throughout the United States. He
went on to say,

"[We] realized also that proper provisions should be made in the

proposals and contracts, whereby adjustments in the contractors'
compensation because of changes in plans could be on a basis agreed upon
in the contracts .. .'>')

The contractors examined Strauss' Invitations for Bids in one of the

most trying times of United States history. By the summer of 1931, the
depression weighed heavily over America. Things were clearly going to get
worse before they got better. The bids, returned in July 1931, were as
shown in Table 1-1.10

I-A Steel Superstructure $10,494,000.00

I-B Steel Cables, Suspenders & Accessories $6,255,767.65
II San Francisco Pier & Fender and Marin Pier $2,260,000.00
III Anchorages & Piers of Approach Spans $1,645,841.28
IV Steel for S.F. and Marin Approaches $996,000.00
V Presidio Approach Road $966,180.00
VI Sausalito Approach Road $67,586.00
VII Paving of Main Span, Side & Approach Spans $345,000.00
VIII Electrical Work $133,495.00
IX Toll Houses & Service Buildings $71,430.00
X Cement (Est. 500,000 bbls.) $1,220,000.00
Total $24,455,299.93

Table 1-1 Bids Returned, July 1931

 Chapter 1 Golden Gate Bridge 7

The bids indicated that if Strauss' design documents remained stable,

the costs of construction alone would come in under not only the bond issue,
but also under Mr. Strauss' estimate of $27 million. When the engineering
fees, legal fees, and interest were included, total project costs were much
closer to $35 million.

Bond Financing for the Project

The six-year fight over the legality of the District proved only to be a
prelude to the opposition's efforts to stop the bridge. As outlined in the
Bridge and Highway District Act, the District had the right to issue bonds,
subject to approval by vote in the represented counties. Through numerous
articles and advertisements in the local newspapers, the opposition
attempted to dissuade the public from supporting the bond issue. The
arguments included those used in the legal fights against the formation of
the District. The press was sprinkled with new insinuations that the
Directors of the District and the Chief Engineer (Strauss) were going to
profit personally from the project. I I (See Exhibits 1-4 through 1-6 on
companion CD for information on the bridge bonds)

Exhibit 1-4 Newspaper Advertisement on the Bridge Bond Issue

Exhibit 1-5 Bond Sales Record

Exhibit 1-5 By Permission, Golden Gate Bridge and Highway District, from The
Golden Gate Bridge, Report of the Chief Engineer, 50th Anniversary Edition, 1987.

Exhibit 1-6 Bond Payment Schedule

Exhibit 1-6 By Permission, Golden Gate Bridge and Highway District, from The
Golden Gate Bridge, Report of the Chief Engineer, 50th Anniversary Edition, 1987.
After the District obtained voter approval for issuing the bonds, one of
the contractors and a group of proposed investors raised questions regarding
the constitutionality of the District's right to tax. These questions alarmed
the financial community since the creditworthiness of the District and its
bonds depended on the security pledged by taxpayers in the member
counties. To solve this new problem, the Supreme Court of California was
asked to decide this question before the bonds were issued and the contracts
8 Case Studies in Infrastructure Delivery

let. The Justices, noting the urgency of the matter, heard the case quite
quickly and, with one Justice dissenting, ruled for the District.

Still, the legal struggle was not over. Yet another suit, by two
companies from the northern parts of California, was brought against the
District, challenging the right of the District to tax for the purpose of
servicing the repayment of principal and interest on the bonds. Garland Co.
v. Filmer, et al. (1932 N.D.Cal.) 1 F.Supp 8. The District won the case at
the trial level and an appeal was withdrawn in 1932.

This seemed to be the last legal hurdle left to issue the bonds and
construct the bridge. But another difficulty arose as steps were begun to let
the contracts and to begin construction. Bankamerica Company, the lead in
a syndicate organized to finance the bridge, bid 92.3 cents on the dollar for
the 4.75% bonds that were to be sold by the District to finance the bridge.
The District's authorizing statute allowed only a 5% interest rate on the
bonds. Even though the bonds were issued at 4.75%, Bankamerica's offer at
92.3 cents pushed the effective yield to 5.25%. Bond attorneys from New
York, retained by the District, thought that the Bankamerica offer, if
accepted, would make the bonds illegal. Counsel local to San Francisco
disagreed, but the District decided that the issue required attention.
Unfortunately there were no funds left for the counsel to bring the matter
before the appropriate court for a decision.

Strauss approached the chairman of the Bank of America, A. P.

Giannini and, with the help of the most powerful people associated with the
bridge's interests, convinced Giannini to form a new syndicate for the
purpose of buying Golden Gate Bridge Bonds. The syndicate was formed
and bravely, considering the depression bond market, agreed to buy
$3,000,000 worth of bonds from the District at 96.23 cents on the dollar,
enough to put the effective interest rate just under 5%. The newly formed
syndicate agreed to forward $184,600 against $200,000 worth of the
$3,000,000, pending the outcome of the legal question regarding the
effective percentage rate. Funding for the preliminary planning and early
design was now available and work continued. This last difficulty was
resolved and the District was finally in a position to proceed with
construction.
 Chapter 1 Golden Gate Bridge 9

Construction

Construction of the bridge involved hundreds of contractors and

subcontractors and many bridge-specific castings and forgings. There were
many opportunities for the potential problems raised by the bridge
opponents to come true and for expenses to run out of control. Despite
several delays and cost overruns, the bridge remained within the first
estimate. Some costs were also saved when part of the San Francisco
approach was undertaken as a city project. Responsibility for that approach
was shifted when the city objected to the planned connection at Marina
Boulevard and insisted that the road connect with Lombard Street. In
addition, responsibility for a portion of the approach on the Marin side was
shifted to the state government. Both of these changes helped The District
keep their costs for the bridge under the $35 million ceiling.

In many cases, construction techniques needed to be altered or invented

anew to meet the needs of this bridge. The San Francisco-side foundation
pier, built essentially in the open ocean, required an access trestle 1100 feet
long and 22 feet wide running out into the mouth of the Golden Gate. The
trestle was rammed by an off course vessel, tearing out a section of the
structure and seriously damaging the ship. The trestle, seriously weakened
by the encounter with the ship, was further damaged in a severe storm and
lost over 2/3 of its length into the ocean. Rebuilt, the trestle remained for
the duration of construction, allowing access to the San Francisco pier via a
newly secured sub-structure tied to the bedrock with steel cables.

Plans for the San Francisco pier itself underwent four expensive
revisions before a final construction approach was chosen. The original
plan was to pour a concrete fender that would protect an oval section of
water out in the bay. Within that fender, a pneumatic caisson would be used
to reach dry ground at the bottom of the bay. Construction would then
progress as if the tower pier were on dry land rather than 1,200 feet out into
the ocean. However, when the fender was partially complete, and the
caisson was floated in from the open eastern half, the wave action inside the
half-enclosed fender was sufficient to nearly destroy the caisson. A quick
decision was made to abandon that method of construction and the caisson
was floated back out without undue damage to the fender. The entire
custom-built pneumatic caisson was later sold for scrap.
10 Case Studies in Infrastructure Delivery

The San Francisco pier eventually cost as much as the estimate for both
piers, and that was but one cost overrun. The repairs and adjustments to the
military reservations, originally estimated at $100,000, eventually ran to
$575,000. Military cooperation was crucial to the project. Continued
approval was required for the access roads, approaches, and tollhouse area.
Every new application for changes in use was met with additional
complications. The military requisitioning procedure meant massive delays
and cost overruns. By the end of construction, The District had paid for the
removal and rebuilding of barracks, roads, sewage and drainage systems,
fire stations, gas stations, a new powder magazine (costing $125,000.00),
machine shops, and a new rifle range. 12

The cable saddles that guide the cables (in an octagonal formation) over
the tops of the towers, the cable bands that clamp down on the main cables
and attach the suspender ropes, and the strand shoes that allow the main
cable strands to attach to the anchorages were all custom castings. Each one
was cast and machined to a smooth finish, allowing even bearing between
the cable and casting. These were all subject to testing for cracks and voids.
The cable bands in particular were notorious for voiding (hardening with air
pockets deep within the metal) and had to be re-cast many times over. The
errors in casting were only evident when the part was finished in the
machine shop, but at that point the expensive work had already been done.
The cable bands, and the rest of the castings were very expensive.

During construction, battles of public opinion continued to be fought in

the local papers. Once, during the construction of the San Francisco pier, a
local geologist questioned the seven experts' opinion regarding the safety of
the founding of the pier. The directors of The District were brought by the
tide of public fear to a public forum where they refuted the claims of the
geologist. Early on, the Chief Engineer, Joseph Strauss, formed an
information campaign that remained active from before the formation of
The District until the bridge was completed.

Exhibit 1-7 By Permission, Golden Gate Bridge and Highway District, from The
Golden Gate Bridge, Report of the Chief Engineer, 50th Anniversary Edition, 1987.
 Chapter 1 Golden Gate Bridge 11

Questions
For simplicity, ignore the effects of inflation in the following questions.

1. Compute Your Own Estimate of Net Present Value

Familiarize yourself with Exhibit 1-8 on the companion CD. This

worksheet contains the initial assumptions relating to cash flow for the
project prior to the letting of contracts.

Exhibit /-8 Calculations Worksheet

Copy the worksheet into your own file and fill in the balance of the
cash flows. State your assumptions regarding the timing of the bond
sales, the distribution of construction costs, and any other assumptions
you make in filling out the worksheet.

The worksheet includes reduced toll rates over time. This was
included in the project cost and revenue estimate. Exhibit 1-9 shows
spikes in toll receipts over time (the top curve). Ifrates are going down
and revenues are reasonably flat, what does this tell you about the
projected traffic volume?

Calculate the NPV of the cash flow you generate.

Exhibit /-9 Forecast o/bridge traffic and revenue

Exhibit 1-9 By Permission, Golden Gate Bridge and Highway District, from The
Golden Gate Bridge, Report of the Chief Engineer, 50th Anniversary Edition, 1987.
2. Sensitivity to Changes in Assumptions

Key assumptions made by Strauss regarding toll pricing eventually

proved to be wrong. You are not surprised, are you? For example, in
response to the completion of the Bridge, in 1936, the ferries dropped
their prices. The Golden Gate Bridge Authority was forced to adopt a
lower toll rate than assumed in the forecasts.
12 Case Studies in Infrastructure Delivery

Assumptions used in any financial forecast are just that. Question

2 tests the sensitivity of the NPV you calculated in Question 1.

Compute the NPV of the cash flow for each of the following cases,
and compare it to the base case you calculated in Question 1.

a. A 10% increase in toll paying vehicles across the entire cash

flow.

b. A 10% decrease in toll paying vehicles across the entire cash

flow.

c. A delay in construction completion of 2 years (at the same total

cost).

d. An increase in the total construction price of 10% (allotted in the

same proportions as your base case from Question 1).

e. A 25% increase in the planned toll rates.

f. A 50% increase in the planned toll rates.

3. Answer the Following Questions.

a. What was the project delivery method for the Golden Gate
Bridge? Did the Counties "outsource" the project's cash flow
(sources and uses) to an independent entity? Why? Is the
Authority a public entity? A private entity?

b. In which quadrant does the Golden Gate procurement fit?

c. Some components of the project were not completed within the

original estimate. For example, the San Francisco pier
experienced a 25% overrun. The military refurbishments were
nearly 500% over budget. What degree of cost overrun for the
whole project would have caused the "failure" of the project?
(How much cost overrun would cause the project to "fail"?)
 Chapter 1 Golden Gate Bridge 13

d. Put yourself in the shoes of Bank of America. Why would you

agree to buy the bonds at such a rate? What were your Risks?
Potential Rewards? Does your NPV analysis help justify these
risks?

e. Based on the NPV results, did citizens have to worry about the
Authority's ability to repay the bonds? Were the tolls capable of
handling debt service payments?

References
Cassady, S. Spanning the Gate. Mill Valley, California: Squarebooks, Inc.,
1979.

Golden Gate Bridge and Highway District. Annual Reports 1937/38 -

1979/80. San Francisco, CA: Golden Gate Bridge and Highway
District, 1938-1980.

Newspaper Scrap-Book. (Articles principally from the San Francisco

Examiner and the Santa Rosa Independent). Property of The Golden
Gate Bridge and Highway District, San Francisco, CA, ca. 1928 - 1937.

Strauss, J. and C. Paine. The Golden Gate Bridge: Report of the Chief
Engineer to the Board of Directors of the Golden Gate Bridge and
Highway District, California. San Francisco, CA: Golden Gate Bridge
and Highway District, 1938.

Notes

I Report of the Chief Engineer, J. B. Strauss, p. 25

2 Report of the Chief Engineer, J. B. Strauss, p. 27
3 Spanning the Gate, S. Cassady, p. 43
4 Report ofthe Chief Engineer, J. B. Strauss, p. 27
14 Case Studies in Infrastructure Delivery

5 The group that applied to Colonel Deakyne did so based on the supposition that an
official district would be formed to administer and finance the bridge. That district was
made possible by the Bridge and Highway District Act of California, May 25, 1923.
6 That provisional permit eventually saved the project in 1930, when final approval was
sought before General Lytle Brown, who opposed the bridge but could not rescind the
provisional permit presumably because to do so would dishonor General Deakyne's
judgment (Report of the Chief Engineer, l B. Strauss, pp. 31 & 39). It is difficult to
imagine that the same line of thinking would be allowed to prevail in today's
environment.
7 Report of the Chief Engineer, l B. Strauss, p. 42.
8 This illustrates an early of the Design-Bid-Build concept. Clearly Mr. Strauss
identifies standard practice as evolutionary, where units are bid in sequence, as they
become needed. This probably allowed for a greater degree of flexibility in design and
construction, but greater uncertainty in cost.
9 Report of the Chief Engineer, lB. Strauss, p. 44. Note that this is an early use of a
changes clause on a major infrastructure project.
10 Report ofthe Chief Engineer, J. B. Strauss, p. 44
11 Most signature projects like the Golden Gate Bridge produce similar allegations-
some real, some imagined. The Brooklyn Bridge, the New York Subway, the Eads
Bridge, and the Central Artery Project in Boston are examples.
12 Report of the Chief Engineer, J. B. Strauss, p. 51