Toward a Portfolio Based Procurement Strategy for Multiple Projects

Part of the The Springer International Series in Infrastructure Systems: Delivery and Finance book series (ISDF, volume 101)


Chapter 6 looks to the future, a future in which direct and indirect project delivery and finance methods are simultaneously available to governments and to private sector owners of infrastructure collections.1 The focus now turns to very practical questions: how to configure portfolios of infrastructure projects so that different mixes of projects, delivery methods, and financing approaches can be evaluated to positively change portfolio quality, revenues, expense, and technology. Political bodies, elected officials, and public officials will not use the delivery methods of Quadrants I, II, and IV unless they are reasonably sure that their choice of delivery method produces real improvement in infrastructure quality, level of service, cost performance, or timeliness. Similarly, corporate officers, corporate infrastructure managers, and corporate boards of directors need to know, not guess, that proposed infrastructure expenditures will produce incremental improvement in corporate infrastructure performance.

Chapters 2 through 5 demonstrate that choice of delivery method, source of financing (owner, producer, users, or combinations thereof), level of investment (“pace”), and level of service (cost and quality) are variables that public and private CLIENTS will continue to use across an infrastructure portfolio. “Optimization” is not a viable approach2 because many of the variables that decision-makers apply can only be subjectively measured. CLIENTS will find themselves selecting one among many configurations for the portfolio, each of which attempts to maximize portfolio performance while using all available financial resources.

Chapter 6 begins with a look at current budget and accounting practices in several small communities in New England. The public setting is chosen because data is not only easy to acquire, but the data itself is public information. In each of the jurisdictions described in this chapter, historical sources and uses of all funds available to each community were tracked in a database generated from audited financial statements and locally adopted budgets. All the data collected complied with standard accounting and budgeting practices. Each of these communities is well run, by highly motivated elected and appointed officials.

The recorded financial data is detailed and accurate. Yet, this data is not very useful: (a) for projecting future repair and replacement obligations, (b) for establishing the current condition of existing infrastructure assets, or (c) for making reliable predictions of the effect of an infrastructure project on the sources and uses of funds for the entire portfolio of infrastructure assets. Each of these three deficiencies puts practical pressure on public and private sector decision-makers not to use the project delivery alternatives strategically, that is, for the overall gain of the entire portfolio.3 The most visible effect of the current deficiencies in public financial data (and that of many private firms) is the confusion generated from abstract and futile arguments over the “best” project delivery and finance method throughout North America.

Chapter 6 explores structural problems that arise from these deficiencies through a more detailed look at capital programming in the City of Medford, Massachusetts. Again, this review is not to criticize, but rather to characterize the difficult challenges facing public CLIENTS trying to manage existing assets and improve and expand future services, all in competition with other government activities and with constrained budgets.4

A new methodology is presented that combines project delivery alternatives with condition assessment and activity cost principles into a single comprehensive approach to infrastructure asset management. The new approach has two key components: the first is a substantial upgrade in the capacity of owners to dynamically describe the current condition of their infrastructure collection.

The second component of the new approach is a scenario-based approach to the simultaneous application of multiple delivery methods across a collection of infrastructure projects. A prototype of this decision support system — called CHOICES, OMIT 1998 and 1999 — has already been built at MIT. The model is positioned so that key attributes of each project in the portfolio — such as start date(s), duration(s), multiple project delivery methods, and various project finance structures — can be quickly input for each project in the portfolio. The analysis can quickly access each of the different feasible delivery strategies for each project, so different scenarios for the entire collection of projects can quickly be assembled. CHOICES allows decision-makers to compare and contrast these scenarios with a very clear understanding of how each scenario allocates constrained resources differently across the entire collection. Scenario analysis allows engineers and planners to use financial constraints as the common denominator to choose one configuration from among many and to put the owner at the resource constraint “most effectively.”

The scenario approach gives public and private owners (legislatures, secretaries of transportation, councils, boards of directors) real choices at the portfolio level. In its simplest form, each and every portfolio configuration is constructed to expend all available resources, i.e. each configuration represents a different combination of projects, start dates, schedule duration, and project delivery and finance structures. All of these configurations are at the “resource frontier” — i.e., there are no further resources available if any of the scenarios are adopted. This puts complex numerical analysis out of the way, so that decision makers have a clear look at how different projects, start dates, delivery methods, and financial structure affect the quality, cost, and timing of infrastructure services throughout the portfolio. With fixed financial constraints, the scenario approach allows owners to focus on picking projects and configuring the portfolio to produce the greatest benefit at the resource constraint. The question is reduced to a simple question. Which combination (or “scenario” or “configuration’) of projects draws the greatest support among decision-makers? Inevitably, the selection of one scenario means that some projects are not authorized, some are deferred, while other projects are accelerated.

In the real world, although financial resources are constrained, we know that some constraints are adjustable.5 Mid-term adjustments in the level of investment in infrastructure are common; indeed, such adjustments are a routine part of capital programming. Even in this dynamic environment, the scenario approach is still preferable as the level of financial investment varies. Now, the decision-making process has two steps. The first step is the same, except that rather than fixing the level of investment, the analysis begins with an assumed level of investment. The second step in the process is to change the initial assumption to different level(s) of investment — either higher or lower, or with a different distribution of investment over time. A second scenario building effort is now possible. For example, how does a change in investment level affect the number, timing, and quality of infrastructure services that will be delivered at the newly assumed financial frontier? New questions emerge, and new scenarios are possible. What would the preferred configuration of projects look like if resources were held constant over the next ten-year period? What would the preferred configuration of projects look like if resources were increased at 3% per year? How would financial constraints change if tax rates, user fees, water rates, sewer rates, or tolls were restructured (up or down)?

The CHOICES methodology was applied on a test basis to the City of Medford, Massachusetts, a small city with infrastructure assets valued at approximately $500 million. The Medford case points to the direction of future research in capital programming, project delivery, and project finance. Condition assessment tools and cost accounting records need to be substantially improved so that repair projects and new projects are always considered together in capital programming scenarios. Repair projects must compete with new projects for scarce financial resources, and the scenario approach allows this competition to be conducted by comparing the effects of any project on current and future cash flow at the “resource frontier.”


Cash Flow Delivery Method Life Cycle Cost General Fund Project Delivery 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    The use of Design-Build, Design-Build-Operate, and Design-Build-Finance Operate as viable project delivery methods was described in Chapters 3, 4, and 5. Flexible integration of the design, construction, finance, and operation functions was a well-established practice in pre-depression America. The current resurgence of these methods in the United States and Canada follows the experience in Hong Kong, and is in parallel with strikingly similar developments in Singapore, Australia, Taiwan, and Korea.Google Scholar
  2. 2.
    Optimization is used here to mean some kind of process (weighted factors, conversion of all decisions to numerical equivalents, or financial representations) that would tell a decision-maker that one particular configuration of all the projects is “optimal.”Google Scholar
  3. 3.
    How can a corporate officer or a public official rationally choose project A (and delivery method 1) over Project B (and delivery method 2) unless the qualitative and financial effects of the choice are reasonably well known and understood! Project A might eliminate a very expensive, ongoing repair obligation, which needs to be included in any comparison with Project B. Project B might involve the demolition of a facility that is in excellent condition, with substantial useful life remaining.Google Scholar
  4. 4.
    Mayor Michael McGlynn, of the City of Medford, provided great assistance to the author and his students in the development of our analysis of cities like Medford. City personnel are highly motivated, well qualified, and performing very capably for the City of Medford.Google Scholar
  5. 5.
    For example, the Loma Prieta earthquake in Northern California changed the level of investment on structural retrofit in California.Google Scholar
  6. 6.
    Direct and Indirect project delivery methods offer distinctly different cash flow results to owners in their efforts to match limited financial resources to specific projects. Segmented and Combined project delivery methods offer distinctly different results too for owners, not only in terms of cash flow but in the nature of the control exercised by an owner over initial costs, life cycle costs, and quality. Both public and private owners can effectively retain control over infrastructure results in all the quadrants, but earlier planning is required to properly manage the combined delivery methods of Quadrants I and II.Google Scholar
  7. 7.
    Many cities and towns in the Metropolitan Boston area have established separate “enterprise” accounts since the establishment of the Massachusetts Water Resources Authority, which now supplies drinking water and treats wastewater for numerous eastern Massachusetts cities and towns on a “wholesale” basis. Winchester is one of many such towns that sets water and sewer rates to exactly match the costs charged by the MWRA and the costs of local distribution and collection.Google Scholar
  8. 8.
    The cap was enacted through a referendum initiative known as Proposition 2 1/2. The cap on real estate tax rates also included a similar cap of 2.5% per year cap on the amount that real estate tax rates could be raised in any one year.Google Scholar
  9. 9.
    The general fund expenditures do not include the Enterprise Funds for water supply and wastewater treatment.Google Scholar
  10. 10.
    The “Closure Line” in Figure 6–3 simply represents the difference between sources and uses of funds in the Town’s financial records for that quarter. These differences may arise from small errors in the documents, variations in the actual amounts of taxes receives, or in the actual amount of expenditures made. The extraordinary amount of time required to tie these small differences down was avoided by establishing the closure line concept.Google Scholar
  11. 11.
    Recommendations of the Capital Planning Committee, January 11, 1993, Town of Winchester Budget documents, FY 94.Google Scholar
  12. 12.
    Recommendations of the Capital Planning Committee, Town of Winchester Budget documents, FY 96.Google Scholar
  13. 13.
    The title to the Town Manager’s Budget for FY99: “Return of HARD DECISIONS - TOUGH CHOICES” confirms our conclusion: striking a balance among competing programs, current repair, and future capital obligations presents nothing but tough choices and hard decisions.Google Scholar
  14. 14.
    Boston Globe, December 15, 1999, p. BI. The article describes the problem.“It was the full-throated cry of Beacon Hill lawmakers and education officials last month: ‘We must keep the promise to our school children.’ The legislators unanimously overrode Governor Paul Cellucci’s veto and gave local schools $94 million on top of the $3.5 billion already appropriated for them.But some of that much-disputed money will not be spent on school children, but on cutting local taxes, building a new police station, or adding to town general funds. Several communities say they didn’t plan on the extra money and had already set final school budgets for the year. Other districts say their schools really didn’t need the cash. For example, Boxborough cut its property tax rate 6 percent after the veto override, which brought $950,000 in unexpected education aidGoogle Scholar
  15. 15.
    An ongoing study of another New England state confirms the insidious effects of year-to-year approvals of capital appropriations on the overall status of infrastructure portfolios. Annual capital requests for preventive repairs and capital replacements exceed annual appropriations granted by the legislature by factors of between ten and twenty. Long term capital planning cannot be conducted in such environments.Google Scholar
  16. 16.
    The intent is not to make judgments. There are certainly no grounds for criticizing any City’s decision to allocate scarce resources to education, police, and fire services.Google Scholar
  17. 17.
    If the City reassessed all real property at 100% of actual value, the ration of net bonded debt to assessed value would drop, but net bonded debt per capita would remain unchanged.Google Scholar
  18. 18.
    Just as in the case of Winchester, the “Closure Line” in Figure 6–7 simply represents the difference between sources and uses of funds in the Town’s financial records for that quarter.Google Scholar
  19. 19.
    We’re not talking about routine repair or operation tasks that may forever be provided by a talented, motivated municipal staff.Google Scholar
  20. 20.
    Mayor McGlynn has been extremely supportive of the notion of capital planning at the portfolio level, for which the author and several MIT graduate students are very grateful. Many hours were spent analyzing data provided by key department heads in the City of Medford, which was crucial in our effort to test CHOICES in a mid-sized city. Michael J. Garvin and Steve Wooldridge, both Ph.D. students in MIT’s construction management program participated in the development of the Medford case study.Google Scholar
  21. 21.
    Enterprise projects (water and wastewater services) were analyzed separately.Google Scholar
  22. 22.
    To keep the example relatively simple, and to explore the possibilities without incurring additional debt, a third variable on the revenue side - municipal debt - was assumed not to be a variable. Figures 6–15 through 6–17 shows municipal debt to be a small contributor to revenues for capital program needs.Google Scholar
  23. 23.
    Aggressive bond financing alternatives were not explored.Google Scholar
  24. 24.
    These reasons are described throughout this book. Infrastructure facilities deteriorate over time, dependent, in part, on the level of preventive maintenance and repair. Technology changes over time, which can rapidly change the efficiency of existing facilities in terms of energy consumption, pollution, cost, and quality. Public demand for infrastructure services change over time, as forms of transportation, communication, and life-styles change.Google Scholar
  25. 25.
    This new paradigm is described by Miller, John B., Michael J. Garvin, P.E., C. William Ibbs, and Stephen E. Mahoney, Towards a New Paradigm: Simultaneous Use of Multiple Project Delivery Methods, Journal of Management in Engineering, ASCE, Vol. 16, Issue 3, 58–67.Google Scholar
  26. 26.
    Cash flow models were introduced in Chapter 2.Google Scholar
  27. 27.
    The term “project” is used broadly throughout the text, but particularly in this chapter. Delivery of a new facility is a “project,” but so is the replacement of a roof on an existing public building. The term “project” is applied in the CHOICES model to all separately budgeted tasks related to the delivery, repair, or operation of infrastructure facilities and services.Google Scholar
  28. 28.
    CHOICES is copyrighted by MIT in 1997, 1998, and 1999, All Rights Reserved.Google Scholar
  29. 29.
    Miller, J. B., and Evje, R. H. “Life Cycle Discounted Project Cash Flows: The Common Denominator in Procurement Strategy.” First International Conference on Construction Industry Development, SingaporeGoogle Scholar
  30. 30.
    The model requires that sources and uses of funds balance in the aggregate and by quarter. To accomplish this result, shortfalls and surpluses are calculated for each quarter and displayed throughout the model. The model displays the quarterly and aggregate status of surpluses and shortfalls for each and every configuration. Analysts can quickly identify the fiscal viability of each configuration, and can begin to consider what, if any, cash management strategies might be required to take advantage of surpluses or to avoid shortfalls.Google Scholar
  31. 31.
    See the discussion of Project 0, in the next section.Google Scholar
  32. 32.
    DBO and DBFO projects are typically 20 — 25 years in duration, a period sufficiently long to period long term debt financing with principal and interest payments that can be supported by project revenues (much like a home mortgage).Google Scholar
  33. 33.
    Brealey, R. A., and Stewart C. Myers. (1996). Principles of Corporate Finance, The McGraw-Hill Companies, Inc., New York.Google Scholar
  34. 34.
    This is why CHOICES is described as a decision support tool. The output of the model can be better than the data entered.Google Scholar
  35. 35.
    Recall that we are defining DBFO (and BOT) to be a Quadrant II process in which the revenue and expense risk associated with a particular project are bundled together by the CLIENT (public or private) and competitively awarded to a private sector PRODUCER. The CLIENT provides no actual cash flow, although it will typically support such projects indirectly, as described in Chapter 3, and in Appendices A and B.Google Scholar
  36. 36.
    Gordon, C. M. (1994). “Choosing Appropriate Construction Contracting Method.” Construction Engineering and Management, 120, No. 1(MAR. 1994), 196–210.Google Scholar
  37. 37.
    In the public sector, public funds for projects are often made conditional on the delivery method used. Design-Build, for example, has been a required delivery method in the Federal Transit Administration’s Turnkey Demonstration Program. Design-Bid-Build is often a federal legislative requirement, which cannot be changed without refusing federal support. An analyst can accommodate such requirements in CHOICES in a number of ways. First, the analyst can limit the preparation of project cash flows to the legislatively prescribed delivery method. Alternatively, the analyst can prepare multiple project cash flows across the delivery methods to permit scenario analysis, but fix the cash flows used in configurations to the prescribed method.Google Scholar
  38. 38.
    Miller, J. B. “Procurement Strategies Which Encourage Innovation: The Fundamental Element of Sustainable Public Infrastructure Systems.” Montreal, Quebec.Google Scholar
  39. 39.
    The Silver Lane would establish a new direct rapid transit connection from each of Boston’s four existing subway lines to Logan International Airport. The portfolio modeled with CHOICES assumes the new line beginning at Boylston Street Station (Green Line connection) and traveling through Chinatown (Orange Line connection) to South Station (Red Line connection) to the World Trade Center in the Piers district, across the Harbor, through the Airport Terminals, and to the Blue Line at Airport Station.Google Scholar
  40. 40.
    Each configuration will have advantages, disadvantages, tight schedule points, tight financial points, and, in cases where costs far outstrip ability to pay, a clear set of legislative or corporate choices as to how to re-configure the portfolio.Google Scholar

Copyright information

© Springer Science+Business Media New York 2000

Authors and Affiliations

  1. 1.Massachusetts Institute of TechnologyCambridgeUSA

Personalised recommendations