Financing

1
Financing

• David Levinson

2
Financing c. 1920
3
Financing c. 1996
4
Federal Highway Revenue by User Class
5
Freight Movement (Share of Ton-km)
6
Federal Transportation Budget
7
Transportation Revenues and Revenue Raising
Instruments by Mode FY 1999
8
Public-Private Partnerships

• Rising infrastructure costs and social demands gt
  tight government budgets and public resistance
  (tax increases)
• Mutually agreed government agencies and private
  business
• Private sectors fulfilling responsibilities

9
PPP Types and Model

• Main Types of PPP (USGAO, 1999)
• Build-Own-Operate (BOO)
• Build-Opeate-Transfer (BOT)
• Buy-Build-Operate (BBO)
• Design-Build-Operate (DBO)
• Build-Develop-Operate (BDO)
• Similarities (European)
• Design-Build-Finance-Maintain (DBFM, The
  Netherlands Case)
• Design-Build-Finance-Operate (DBFO, Ireland)

10
USA PPP Experience

• The United States Large PPP projects in the road
  sector
• Alameda Corridor (Long Beach-Los Angeles, CA)
• Dulles Greenway, Virginia
• State Route 125 South Tollway (San Diego County,
  CA) 4 lane 11.2 mile highway
• Western Loop (Richmond, Virginia)
• Route 3 (Massachusetts)
• Highway E-470 (Aurora, Colorado) 47 mile toll
  road

11
Alameda Corridor

• Alameda Corridor (Long Beach-Los Angeles, CA)
• History
• Ports Advisory Committee (October 1981)
• Route Alternatives (1984)
• Alameda Corridor Transport Authority (1989)
• The US DOT 400 M, 30 year loan (1998)
• Objectives
• Reduce Highway traffic delays
• Increase Rail productivity
• Reduce Accidents
• Improve Air quality

12
Alameda Corridor

• Related Projects
• San Gabriel Valley 55 rail crossings, LA County,
  8 year project, 912 M
• Pacific Coast Highway Separate Grade (PCH)
  30,000 commuters daily, 1 mile long
• Discussion
• Completion on time and on budget
• Design-Build instead of Design-Bid-Build
• Creation of 700,000 jobs (2020, port growth)
• Environmental benefits truck traffic (23
  reduction), delays at grade crossings (90
  reduction), noise and vibrations (90 reduction)

13
Dulles Greenway, Virginia

• History
• Originated 1988 Virginia Assembly authorized
  private toll roads
• Construction initiated in 1993 and concluded in
  1995
• Consists 7 interchanges, 36 bridges, a toll
  plaza, 12 ramp toll barriers, 4 operational
  lanes.
• Extension 14-mile western extension, Dulles Toll
  road.

14
Financing Regulation

• The Case of Spring Load Restrictions.
• Current Truck weights restricted during spring
  thaw
• Alternative Trucks allowed but charged for the
  damages imposed.

15
How Much Revenue?

• The revenue required to compensate road owners
  for the additional damage associated with lifting
  the SLR in Minnesota can be estimated.

16
Estimate of Statewide Annual Cost

• The annualized cost of a net present value of
  127,457,204 at 3.5 interest over 42.5 years is
  therefore 5,806,768.
• Different assumptions will yield different
  annualized costs.

17
Diesel Fuel Surcharge

• Presently the tax on diesel fuel is 0.20/gallon
• In total, 652,549,000 gallons of Special fuels
  were consumed in Minnesota in 2002
• This implies to cover the costs of removing the
  SLR on 7 and 9-ton roads, a year-round
  0.01/gallon diesel fuel surcharge would be
  sufficient

18
Annual Fee

• There were 34,729 truck/tractors and 48,938 farm
  trucks in Minnesota in 2002
• Allocating the cost uniformly to all
  truck/tractors and farm trucks would give a
  charge of 69.40 per farm truck and truck/tractor
  vehicle per year to recover the additional damage
  to roads associated with lifting the SLR on 5 and
  7-ton roads.

19
Weight Distance Tax

• In 1999, Oregon voters passed Measure 76, and
  placed in the state constitution the idea of
  cost responsibility, ensuring that cars and
  trucks each pay their fair share.
• The Oregon Highway Cost Allocation Study is
  conducted biennially to support highway-financing
  decisions.
• The 2003 report states that light vehicles
  (weighing 3,636 kg (8,000 pounds) or less) should
  pay 66.6 of state highway user revenue, and
  heavy vehicles should pay the remaining 33.4.
• Employing a weight-distance tax in Minnesota
  would require a change in revenue policy well
  beyond what is required to recover costs from the
  Spring Load Restrictions, but remains a good idea
  to maximize both fairness and efficiency in the
  highway financing system.

20
Permitting System

• A new SLR permitting system would require a new
  regulatory apparatus.
• Though it would be possible in principle to
  charge directly based on use, the enforcement
  required to do so would entail a significant
  transactions cost that may obviate the gains from
  policy change.

21
How Should the Revenue Be Spent?

• Because most of the economic burden associated
  with lifting the Spring Load Restriction would be
  borne by local governments (counties and
  municipalities), the revenue that is collected to
  recover the costs of the additional pavement
  damage associated with lifting the SLR should be
  dedicated to local governments to spend on
  maintaining and rebuilding roads. Local
  governments would then need to prioritize
  projects based on local engineering and other
  information.

22
Discussion (1)

• The highway system has a disjoint control of
  trucks (owned by trucking firms) and pavements
  (owned by governmental road agencies), which has
  created a number of extra costs that proper
  management of the system might avoid.
• Pavements are rated for different loads of
  trucks roads are restricted to 5-ton, 7-ton,
  9-ton, and 10-ton axle weight trucks.
• Shipments across this network are constrained by
  the lowest weight limit permitted on the roads to
  be used (or risk violation though weight
  enforcement off the interstate highways is very
  sparse).
• Some roads should be upgraded, some trucks should
  have more axles, but the disjoint nature of the
  control makes this coordination difficult.

23
Discussion (2)

• The first solution to these problems lies in
  rethinking highway financing. The ability to
  charge truckers different amounts for different
  roads would put the proper incentives for
  socially beneficial behavior back in the system
• A second solution, to improve materials to the
  point that they are too cheap to meter, that is
  so that they are sufficiently strong that it
  doesnt matter the load using them (within
  reason), is the analog to building your way out
  of congestion. Laying pavements with near zero
  variable (per use) costs may be technically
  possibly, but their upfront fixed (one time)
  costs are likely to be very high.

24
Models

• David Levinson

25

• All forecasts are wrong, some forecasts are more
  wrong than others.

26
Rise of the Automobile/ Highway System

• Do cars and cities mix?
• Can new highways (expressways, freeways) be used
  to reshape cities?
• Original interstate plan proposed to bypass city
  centers. Cities demanded connection.
• Some argued highways would help recentralize
  cities (e.g. Regional Plan for New York), Other
  argued highways could be force for
  decentralization into Garden Cities (e.g.
  Regional Planning Association of America)

27
History of Transportation Models

• The Chicago Area Transportation Study (1955) (Led
  by J. Douglas Carroll)
• The study cost 3.5 million and took seven years
  to complete.
• Considered
• the foundational study of urban transportation
  planning in America (MacDonald 1988 Weiner
  1987)
• a model of "rational planning" model (Black
  1990).

28
Before and After

• Detroit Metropolitan Area Transportation Study
  (1953-1955) (Under J. Douglas Carroll). Mostly
  done by hand.
• Chicago Area Transportation Study (1955)
• Washington Area Traffic Study (1955)
• Baltimore Transportation Study (1957)
• Pittsburgh Area Transportation Study (1958)
• Hartford Area Traffic Study (1958)
• Penn-Jersey Transportation Study (1959)

• FHWA Planpac Mainframe Model (1960s)
• UMTA Urban Transportation Planning System
  Mainframe Model (1960s)
• Merger of two models to UTPS in 1970s
• PC implementation (Tranplan, MinUTP, Emme/2,
  QRSII, Tmodel,Transcad, System2) in 1980s

29
CATS

• Agency of City of Chicago, Cook County, State of
  Illinois, and US Bureau of Public Roads
• Created in 1955 to analyze travel behavior,
  forecast future needs, and develop long range
  plan.
• In peak year (1956) employed 369 people,
  including planners and engineers.
• Followed principles of rational planning
• Used quantitative methods to establish technical
  expertise
• Developed first computer-based regional models

30
Rational Planning Model

• Identify needs
• Set objectives
• Develop options
• Evaluate options
• Select best option
• Implement policy
• Evaluate outcome

31
Critique of rational planning

• Top-down (Newtonian model)
• Identical people/groups (aggregation)
• No externalities
• No dynamics, everything in equilibrium
• Everything is objective
• Missing feedbacks
• Others lt_______gt

32
CATS Goal To secure a transportation system for
the Chicago area which will reduce travel
frictions within the constraints of safety,
economy, and the desirable development of land"

• Objectives
• increasing speed,
• increasing safety,
• lowering operating costs,
• economizing on new construction,
• minimizing disruption, and
• promoting better land development.

33
Inputs and Outputs

• CATS Input Data
• travel,
• land use,
• networks
• desire lines
• vehicle counts
• origin-destination surveys
• home interview surveys

• CATS Forecasted
• population,
• population distribution,
• per capita income,
• auto ownership,
• travel behavior

34
Questions

• Did desire lines reflect desires?
• Is past behavior reflective of future behavior?
• Can the future be predicted
• Is the future independent of decisions, or are
  prophesies self-fulfilling?
• How do we know if forecasts were successful?
  Against what standard are they to be judged?
• What values are embedded in the planning process?
  What happens when values change?

35
Purposes of Modeling

• Estimation in the absence of data
• Forecasting
• Scenario Testing (alternative land uses,
  networks, policies)
• Project Planning/Corridor Studies
• Growth Management/Development Regulation/Public
  Facility Adequacy
• Manage Complexity, when eyeballs are
  insufficient, different people have different
  intuitions
• Understanding travel behavior
• Influence Decisions

36
Outputs from Models

• Flows on links,
• Speeds on links
• Origin Destination pattern
• Mode Split
• Other desired outputs
• Emissions (requires post-processor, knowledge of
  fleet composition, dynamics of speeds)
• Time of day splits
• Change in land use as a result of network

37
Modeling Process

• Specification
• Estimation
• Implementation
• Calibration
• Validation
• Application
• Each step feeds back to previous steps.

38
Four-Step Models
39
Network Framework

• zone centroids - special node, number of a zone,
  identified by x y coordinate
• nodes number identified by X Y coordinate
• links, indexed by from and to nodes (including
  centroid connnectors)
• turns, indexed by at, from, and to nodes
• routes, indexed by a series of nodes (e.g. a bus
  route)
• paths, indexed by a series of nodes from origin
  to destination.
• modes vehicle lines, transit lines.

40
Matrices

• Indexed by Traffic Analysis Zones (including
  External Stations)
• 4 types
• scalar,
• vector (origin),
• vector (destination),
• full (interaction)

41
Scalar Matrix

• Scalar
• For example ms01 price of fuel ( per gallon)
• ms01 1.37

ms01 value
42
Matrix Vector Origin

• Origin (i),

• Example ms01 Households per zone

1 .
2 .
3 .
.
.
I .
1 10
2 17
3 12
.
.
I .
43
Matrix Vector Destination

• Destination (j)
• Example Jobs per zone

1 2 3 . . J
. . . . . .
1 2 3 . . J
18 3 560 . . .
44
Matrix Full

• Combine Origin and Destination matrices
• For instance Zone to Zone Trips (Trip Table)

45
Purpose

• Trips are "produced" at an origin and "attracted"
  to a destination. Trips are categorized by
  Purposes, the activity undertaken at a
  destination location
• Typical purposes are
• Home
• Work
• Shop
• School
• Eat Out,
• Social/Recreational
• Medical
• Banking
• Other
• Often categories are dropped and lumped into the
  catchall Other

46
Politics of Modeling

• Try to be neutral arbiter to maintain credibility
  for future applications
• Influence is maximized when the only game in town
  ... try to avoid dueling models
• Modeling is a process not a project, responses
  must be timely, which requires having the model
  set up to answer questions, not setting it up
  after the question is asked
• Spin your own results, don't just dump numbers on
  someone's lap, give the interpretation yourself.
• Only model when necessary, avoid the problem that
  if your only tool is a hammer everything looks
  like a nail.

47
Future of Modeling

• Transsims - activity based ... follow individuals
  rather than aggregates
• Uses simulation and stochastic distributions
• Problems
• Data needs are huge
• computation intensity
• Complexity
• Accuracy
• Still no good answer for trip distribution (which
  requires job matching to be disaggregate)
• Modeling imperfect information, especially
  routing
• Integration of components
• Land-use transportation models

48
Land Use Models

• Objective To predict where new land uses will
  occur, their density, number of units, etc.
• Land Use f (Accessibility - from travel demand
  model, other things)

49
The Seven Deadly Sins of Models (Lees Requiem)

• 1) Hypercomprehensiveness Meaning that the
  models tried to replicate too complex a system in
  a single shot, and were expected to serve too
  many different purposes at the same time.
• 2) Grossness In a way, the converse of
  hypercomprehensiveness. Even though they tried
  to do too much and serve too many purposes, their
  results or outputs were too coarse and
  aggregate, too simplistic to be useful for
  complicated and sophisticated policy
  requirements.
• 3) Data Hungriness Even to produce, gross
  outputs (a few variables), the models required us
  to input many variables for many geographic
  units, and from at least several time periods in
  order to produce approximate projections, and
  very often we could not afford the data
  collection efforts needed to run the models. In
  other instances, data simply didn't exist at the
  levels of specificity which would be
  appropriate to run them.
• 4) Wrongheadedness Lee meant that the models
  suffered from substantial and largely
  unrecognized deviations between the behavior
  claimed for them and the variables and equations
  which actually determined their behavior. As an
  example, when regional averages were used to
  calibrate models, but forecasts were made for
  local areas, the models deviated from reality
  because of specification errors which were often
  not even recognized by their users.

50
Seven Deadly Sins (continued)

• 5) Complicatedness Even though when you looked
  at them through one set of lenses the models
  seemed terribly simplistic, when looked at
  through another set of lenses they were
  outrageously complex. Too simplistic in
  replicating urban economic and social processes,
  the models were too complex in their
  computational algorithms. Errors were multiplied
  because there were so many equations, spatial
  units, and time periods. Even the theoretical
  notion of the model or its representation of an
  urban process was grossly simplistic compared
  with reality. Often, the user didn't know how
  the errors were propagated through series of
  sequential operations and sometimes we needed to
  use systematic adjustments or "correction
  factors" to make the models more realistic even
  though we did not completely comprehend the
  sources of all the errors and could not interpret
  the correction factors in real-world terms.
• 6) Mechanicalness Lee meant that we routinely
  went through many steps in a modeling process
  without completely understanding why we did so,
  and without fully comprehending the consequences
  in terms of validity or error magnification. He
  stated, for example, that even rounding errors
  could be compounded beyond reasonable bounds by
  mechanical steps taken to calibrate and apply
  many models without the user's knowledge.
• 7) Expensiveness The costs of the models,
  derived from their grossness, data hungriness,
  complicatedness, and so on, placed them beyond
  the financial means of many agencies, or depleted
  the resources of agencies so much that the very
  use of models precluded having the resources
  available to improve them or to fine tune them to
  make them appropriate to their applications.

51
Seven Challenges for Land Use Models (Landis)

• 1. Models - microbehavioral (actors and
  agents)Social Benefit/Social Action
• 2. Simulation - multiple movies/scenarios
• 3. Respond to constraints and investments
• 4. Nonlinearity - path dependence in
  non-artifactual way (structure and outcomes,
  network effects)
• 5. spatial vs. real autocorrelation, emergence -
  new dynamics, threshold network effects
• 6. preference utility diversity and change over
  time
• 7. Useful beyond calibration periods. Embed
  innovators and norming agents. Strategic and
  response function.

52
Four Ways to Improve Models (Lee)

• Models should be made more transparent to users
  and policymakers.
• Models should combine strong theoretical
  foundations, objective information, and
  wisdom or good judgment. Without these elements,
  they remain exercises in empty-headed empiricism,
  abstract theorizing, or false consciousness of
  what is actually going on in our urban areas.
• We should start with problems and match our
  methods to the needs of particular situations,
  gathering no more information and using no more
  modeling complexity than is really needed.
• We should build the simplest models possible,
  since complex models do not work well, and
  certainly are unlikely to be understood by those
  who are asked to act on the basis of the model
  outputs.

53
QUESTIONS?

• Additional Slides Follow

54
Definitions

• Speed (V) Distance per Unit Time (e.g. MPH) -
  often called Velocity, measured over space and
  time.
• Flow (Q) Vehicles per Unit Time (e.g. Vehicles
  per hour) - often called Volume, measured at a
  location over time
• Density (K) Vehicles per Unit Distance (e.g.
  Vehicles per Miles) - often called concentration,
  measured over a segment instantaneously
• Detector Occupancy (k) - Percent of time a
  detector is occupied - convert to density if
  vehicle length is known.

55
Fundamental Diagram
56
Space-Time Vehicle Trajectories
57
Queueing
58
Trip Generation Production and Attraction

• The number of trips produced or attracted to a
  purpose in a zone are described by trip rates (a
  cross-classification by age or demographics is
  often used) or equations.
• For instance trips produced from or attracted to
  homes in a zone is described as a function of
• Th f( housing units, household size, age,
  income, accessibility, vehicle ownership).
• From or to work
• Tw f ( jobs(square feet of space by type,
  occupancy rate))
• From or to shop
• Ts f (number of retails workers, type of
  retail, square foot, location, competition)
• Clearly accessibility and vehicle ownership
  require knowing something about the network, and
  so may have to be solved recursively

59
Trip Generation Balancing

• The number of trips produced (at home) from home
  to work must equal the number of trips attracted
  (at work). Two distinct models may give two
  results. Either assume one model or the other is
  correct and adjust the second, or split the
  difference.

60
Activity Analysis

• Frequency- How many times the trip is made per
  day
• Scheduling-the order in which the trips are made
• Activities-home, work, shop other
  (Non-Discretionary).
• Schools, church, visit friends,
  recreation, visit doctor (Discretionary).
• Patterns HWH, HWSH, and HWHSH.
• These are a function of sex, age, employment,
  status, income, auto availability.
• Important things to note in household study are
  the household size (more predictable), household
  structure (less predictable).
• Location/accessibility studies involve feedback.
• Dwelling unit types are obtained from the land
  use pattern and are an indicator of the income,
  race, household structure. They are single units
  and multi-family types.
• Time of day The time of day can be derived from
  the pattern and duration of activities.
  Scheduling models give the pattern of activities
  and not how long each activity takes place.
• In a trip generation framework the peak hour
  factor used is a constant and is a function of
  congestion.

61
Trip Distribution

• Estimate the number of trips going from zone i to
  zone j for each purpose. This requires the
  travel time (and cost) between zones (Cij) and
  the trips produced or attracted to each zone
  e(.g. Th(i), Tw(j)).

Destination Origin 1 2 3 . . J
1 Cij
2
3


I
62
Trip DistributionImpedances

• This table of impedances requires knowing the
  congested travel time on the network, which
  itself requires knowing demand, and so may
  require a recursive solution method of some kind
  (i.e. feedback).
• Interaction between zones is often described by a
  gravity model, in analogy to Newton's Laws of
  Gravitation. While 1/Cij2 was used in the past,
  now a negative exponential form is preferred.

63
Critique of Gravity Model

• Meyer and Miller (1984) claim a "lack of credible
  theoretical basis" for the gravity model, on the
  other hand, it is simply the law of demand, as
  the cost of interaction increases, the level of
  interaction decreases. The only issues are
• (1) What is the shape of the curve
• (2) What else effects demand (i.e. the model is
  incomplete).

64
Negative Exponential Form
65
Opportunities
66
Resulting Trip Distribution
67
Entropy Maximization
Typically Trips can be represented as a function
of productions (P), attractions (A) and
Costs/Times (C) such as Tij f(Aj, Pi, Cij)
Solve iteratively for Ki and Kj
68
Mode Choice

• Estimate the number of trips from each zone to
  each zone by purpose that take mode m.

Where, Cmij is the generalized cost containing
the fares, waiting time, parking cost, transit
environment access etc. Um is the utility of that
particular mode to a person.
69
Variables in Mode Choice Models

• Travel Time of trip
• Travel time to access mode
• Wait Time f(headways of transit vehicles)
• Transfer Time
• Fare
• Parking Costs
• Tolls
• Alternative Specific Constant
• Other Qualitative Data (Sidewalks, Bus Shelters)

70
Relationship of Logit and Gravity

• The observant student will note that the
  functional relationship between the modern
  gravity model and the logit mode choice model are
  very similar, enabling simultaneous choice models
  to be easily developed. The key difference is
  that the gravity model is typically much more
  aggregate.

71
Independence of Irrelevant Alternatives

• Property of Logit (but not all Discrete Choice
  models)
• If you add a mode, it will draw from present
  modes in proportion to their existing shares.

72
Route Assignment

• Auto Assignment This is nothing but how
  travelers choose to go from A to B. Traffic is a
  very dynamic phenomenon.
• Wardorps Users Equilibrium Principle Each user
  acts to minimize his/her own cost, subject to
  every other doing the same Travel times are
  equal on all used routes and lower than on any
  unused route. What people choose is efficient for
  them which however need not be efficient for the
  network.
• Wardorps System optimal Principle Each user
  acts to minimize the total travel time on the
  system.
• Volume Delay Function (VDF) As the traffic flow
  on the link increases, the travel time on the
  link increases. The cost that a driver imposes on
  others is called the marginal cost. The travel
  time for the other drivers increases because of a
  particular driver. While dealing with traffic
  assignment we deal with average cost but while
  dealing with pricing etc. we consider marginal
  cost.

73
Conservation of Flow

• An important factor in road assignment factor is
  the conservation of flow. This means that the
  number of vehicles entering the intersection
  equals the number of vehicles exiting the
  intersection for a given period of time. (except
  for sources and sinks)
• Similarly, the number of vehicles entering the
  back of the link equals the number exiting the
  front.

74
Simple Network
75
Volume-Delay Functions

• Each link has a link volume delay function
  relating travel time on that link and total flow
  on the link. This is analogous to an average
  cost curve used in economics.