Sustainability and Transportation Indicators and Assessment Methodologies
Sustainability and Transportation Indicators and Assessment Methodologies
Sustainability and Transportation Indicators and Assessment Methodologies
METHODOLOGIES
Background Paper for Plenary Session 4
At the
Biannual Conference and Exhibit of the Clean Air Initiative for Latin American Cities:
Sustainable Transport: Linkages to Mitigate Climate Change and Improve Air Quality
25-27 July, 2006
So Paulo, Brazil
Christopher Zegras
Assistant Professor of Transportation and Urban Planning
Department of Urban Studies and Planning
Massachusetts Institute of Technology
77 Massachusetts Avenue, Room 10-403
Cambridge, MA 02139
Tel: 617 452 2433
Fax: 617 258 8081
[email protected]
Of course, many of the problems that people today associate with threats to the
sustainability of modern transportation air pollution, traffic safety, sprawling urban
development patterns, automobile dependence, etc. have been recognized for 50 years or more
(CARB, 2004; Pushkarev and Zupan, 1977; Weaver, 1965). The word sustainable transportation
as understood in the post-Limits to Growth context emerges in the immediate wake of the
Brundtland report, with Replogles (1987) paper at the 1988 Annual Meeting of the
Transportation Research Board on sustainable transportation strategies for the developing
world, explicitly making the link between transportation, basic human needs, and environmental
effects. Since then, we have seen an ever-increasing number of efforts searching to define, design,
and measure sustainable transport (e.g., UN DSD, 1992; OECD, 1996; World Bank, 1996;
WBCSD, 2001; Kennedy et al, 2005; Goldman & Gorham, 2006; etc.). Today one would be hardpressed to find a transportation research project, conference, or increasingly even business
propaganda that, in one form or another, did not include the word sustainable. But, while many
of us now use the word almost reflexively, what does it really mean: lower emissions? of local or
global pollutants? Lower congestion? Higher accessibility? Fewer accidents? Decreased spending
on transport costs? Does it mean all of these? Some of these? A combination of these, and if so,
in what doses? How can we know we are being sustainable and, then, what can we do about it?
A great degree of complementarity (e.g., nearly all comprehensive efforts refer in one
form or another to the so-called Three-Es economy, environment, equity) tends to exist
among the wide-ranging sustainable transport explorations/efforts. But, actual definitions tend to
vary and few if any, operational definitions are proposed. The OECD (2002), for example,
specifically defines a sustainable transport system based on fulfillment of WHO guidelines for air
pollution, noise levels, acidification, and eutrophication, as well as general international goals
related to climate change and stratospheric ozone depletion. Schippers (1996) proposal, that
transportation is sustainable when the beneficiaries pay their full social costs, including those
paid by future generations offers, theoretically, an operational definition (and fully in-line with
the full cost school of transportation analysis, particularly en vogue in the 1990s), yet the
challenges to implementing such a measurement approach are more than daunting.
2.1. Sustainable To Whom and For Whom?
A main challenge to operationalizing sustainable transport comes from the fact that we are
dealing with resource constraints over multiple time horizons with uncertain impacts;
furthermore, we want to ensure that future generations have the same benefits from transportation
as we do; while also ensuring some fair distribution of benefits today. Asking for some trade-off
in inter-generational equity becomes particularly challenging in the developing countries, where
2
sustainability literally is a day-by-day reality for millions of people living on less than one
dollar a day makes it difficult to concern oneself with possible effects tomorrow.
Figure 1. Stylized Representation of a Persons Values Today as they Relate to
Transportation and Sustainability and the Role of A Theoretical Discount Rate
"Worried About Tomorrow"
Accidents
(death and
injury)
Time loss
Short-Term Impacts
Pollution
(acute
effects)
Pollution
(chronic
effects)
Opportunity
cost of land
Vehicle
Disposal
Medium-Term Impacts
Potential
resource
supply
shocks
(fuels,
materials)
Capital
expenses
beyond
revenues
"Permanent" Greenhouse
Land Loss
Gas
Emissions
Long-Term Impacts
Influence
Sustainability
Influence
Articulated
through
Monitored
through
May
influence
Goals &
Objectives
System
Operations
Help define
Evaluation
Criteria
Closely
link with
Indicators
Help
defin
Require
Alternative
Strategies
Define
Requirements for
Data
Used by
Evaluate
d with
Aid in comparison
of
Analytical
methods
In such a planning approach, indicators, quite logically, tie closely to project evaluation criteria
As we would expect, in the face of the boundary discussion above, indicators will vary depending
on the spatial and temporal scale of the analysis and on the ultimate goals, although common
indicators can often apply to several different goals and/or scales of analysis.
Figure 3. The Information Hierarchy through the Sustainable Indicator Prism
(2002) derive a corridor-level index (using multi-attribute utility theory [MAUT] and including
travel time, travel rate, LOS, local pollutant emissions, noise levels, and fuel consumption); Black
(2000) attempts to derive an international-comparative index (based on fossil fuel dependence, air
emissions impacts, traffic accidents, and congestion effects); Yevdokimov (2004) proposes to a
national-level index using the Genuine Progress Indicator (GPI); and several efforts aim to apply
the ecological footprint approach (e.g., Barrett & Scott, 2003).
Sustainability
Dimension
Air Pollution
Environmental
Indicators
Consumption of
Natural
Resources
Health
Social
Indicators
Equity
Opportunities
Economic
Indicators
Costs/Benefits
By Type
Overall
Indicators
Telecommunications
Accessibility
Measure Type
Infrastructurebased
Location-based
Person-based
Utility-based
If accessibility is the ultimate benefit we derive from the transport system (actually, the land usetransport system), sustainability makes us ask: at what cost or what rate of capital drain? How
efficient is our accessibility provision? Some of the costs (infrastructure, vehicles, fares, etc.)
are fairly straightforward they are expenditures that deplete our public and private financial
capital stocks. The other capital drains of, e.g., natural stocks (airsheds, watersheds, etc.), can
be estimated with varying degrees of reliability. But, out indicators, ultimately, need to return to a
common denominator: accessibility, and its efficiency. For example, the key indicator for
transportations carbon efficiency should not be carbon/km traveled; rather, it should be carbon
per accessibility-derived (or, at its simplest, per trip).
3.2. Transportation Analysis: the State of the Art
If we can derive valid and reliable indicators of sustainable transport, can we predict with
adequate confidence the impacts of interventions? Influencing factors in any relevant analysis
of sustainable transport include:
Scale (e.g. metropolitan, intrametropolitan, neighborhood, site-specific);
Scope (e.g. air quality, energy consumption, housing);
Timeframe (e.g. short, medium, long-term).
In rigor, the sustainability concept should cover all scales and scopes over the long-term; the very
word implies a long-term view (i.e., can accessibility be sustained for present and future
generations?). In practical terms, however, we obviously cannot consider all impacts, at all
scales, for eternity. And, in many cases, e.g. a fuel switch or vehicle technology transfer, the
scale, scope and timeframe are basically determined by the technology. The requisite analysis, in
such a case is also fairly straightforward (note, however, that lack of empirical evidence can still
pose a major challenge; Browne et al (2005), for example, show the wide variation in estimates
for hybrid bus fuel consumption rates).
With larger-scale interventions, involving broader transportation system changes, the
analyses inevitably grow in complexity. Such analyses have been the bread and butter of
transportation analysts for at least half a century. Today, the state of the art in transportation-land
use analysis includes simulation models of the integrated decision processes related to land
development, household/firm location choice, household vehicle choice decisions, and
household/individual and firm activities, and finally household/individual and firm travel
activities. With these models, transportation interventions can be analyzed (i.e., alternatives
analysis) and, ultimately, assessment of benefits and costs made. In practice, there are several
difficulties. The state-of-the-art remains fairly distant from the state-of-the-practice; meaning that
the great majority of places use fairly straightforward analytical models (if any), most of these
models virtually ignore commercial/freight transportation, few integrated land use-transportation
analyses are carried out and, finally, the evaluation procedures (to assess the benefits and costs of
alternatives) are not fully capable of analyzing the broad range of relevant impacts. In addition,
and perhaps most importantly, the impact of these techniques (whether state-of-the-art, or not) on
implementation (i.e., decisions made and acted upon) remains fairly case-by-case and, inevitably,
strongly influenced by the political process.
In practice, an interesting question arises regarding the potential to standardize relevant
analytical procedures. In short, can we? It is not clear the degree to which procedures can be
standardized, as the scale of the analytical effort should be matched to the scale of the problem
context (setting) and problem(s). We can see several challenges to effective application of
appropriate techniques: lack of information/data (on passenger/freight transportation behaviors);
and lack of technical capacities, time and resources to carry out the requisite analyses. However,
one could imagine a matrix providing guidelines for the type of analysis, depending on, for
example, the scale, scope, timeframe dimensions outlined above.
At a bare minimum, any relevant analysis should be required to explicitly identify the
analytical boundaries (i.e., Section 2.2), and possible effects on impacts of interest.
3.3. Accessibility and the State of the Art
In practice, it appears that few authorities use accessibility as a performance metric for
transportation systems. Bhat et al (2000) found limited examples of practical use by government
agencies of accessibility as transportation performance measures. The UK Governments
10
Guidance on the Methodology for Multi-Modal Studies (GOMMMS), issued in 2000, includes 3
categories within its accessibility category. The relevant recommendations suggest qualitative
assessment criteria for these categories, considering (rightly or wrongly) that cost benefit analysis
takes into account most aspects of accessibility (UK CFIT, 2004, p. 37).
Hunt (2003) reports on a quasi nested logit model (trip generation-destination choicetime period choice-mode choice-route choice) for the city of Edmonton in which the composite
utility (a measure of accessibility, derived from discrete choice models) fed up to the trip
destination choice provides the measure of aggregate mobility benefits; he demonstrates this
model for estimating GHG emissions reductions and the associated negative effects (decreased
utility) implied. Martnez and Araya (2000) directly link utility-based accessibility measures to
the doubly constrained entropy model (i.e., spatial interaction model); this model is, reportedly,
operational in the Santiago de Chile context.
For a variety of theoretically and practically appealing reasons (truly adherent to the
accessibility-as-benefit for sustainable transportation approach), the future lies in activitybased analyses. Activity-based analyses can allow assessment of travel re-investment and/or reallocation (i.e., induced travel, or longer-trips compensating shorter trips). Operational examples
exist; for example, Dong et al (2005) using the Portland, OR (USA) case estimate an activitybased (as opposed to trip-based) accessibility measure. Despite their theoretical attraction, the
activity-based measures are data hungry and computationally complex.
3.4. Key Methodological Similarities and Differences in Evaluating GHG reductions
and Co-benefits
Many may argue that GHG emissions are the fundamental concern for sustainable transport.
Whether that is truly the case depends, at least in part, on ones perspective (Figure 1).
Fortunately, measuring transport GHG reductions should be entirely compatible with measuring
many other transportation (so-called) co-benefits. After all, fuel consumption, directly related to
GHG emissions is historically a key aspect of traditional transportation evaluation techniques.
But, the similarity depends on the co-benefit of analysis; for example, if available
methodological tools are capable of producing fuel consumption estimates, then local pollutant
estimates should also be derivable. Pollutant emissions estimates have more influencing variables
(cold starts, stops/starts, fuel quality, emissions controls technologies, etc.) and, furthermore, the
need to translate pollutant estimates into ambient concentrations (and ultimately, effects on
health) complicates matters. This remains an area of considerable research activity (e.g.,
incremental v/s. revolutionary improvements to the USEPA Mobile model). While local air
quality analysis has the challenge of the downstream estimation (from the tailpipe to the airshed
11
concentration), GHG emissions face the upstream problem (from source through to end-use
consumption) although we often ignore the latter.
In terms of additional co-benefits (e.g., accidents) within a broader sustainability
framework, these may not always be relevant (e.g., in the case of a basic fuel switch); however, as
the intervention becomes more wide-ranging (e.g., new public transportation infrastructure), the
complexity of analyzing the effects increases. This increased complexity will generally be
reflected in all sustainability dimensions (even fuel consumption calculations will become more
complicated as travel demand and network modeling must be carried out properly). When we
attempt to look at accessibility, distribution of accessibility, the challenges become greater
although, again, a unified and sophisticated forecasting tool (i.e., an integrated land use and
transportation model) should be able to provide the necessary information (accuracy of
predictions of various effects will likely vary). The EU PROSPECTS project (see Minken et al,
2002) should offer some useful guidance on the possibilities.
3.5. Project Counterfactuals and Relevant Assumptions
In the end, when estimating the impacts of particular interventions, we face the challenge of
establishing the baseline. What would have happened without the project. At some broad level,
far into the future, we might be able to use basic mode shares as indicators of potential future. For
example, will Limas non-motorized mode share (in, lets say, 2025) look more like the U.S. or
Germany? What types of policies will influence that outcome (urban development policies,
vehicle and fuel pricing policies, public transport policies, infrastructure investment policies,
etc.)? The current rate of change for example, reported 70% annual growth in Beijings motor
vehicle fleet and concomitant wholesale restructuring of the urban fabric will greatly influence
our willingness to accept the assumption of, e.g., no change in OD patterns.
In terms of the impacts of ambitious proposed schemes (such as large-scale infrastructure
interventions), we must recognize that such projects will likely induce subsequent real estate
investment changes and, quite possible, future additional transportation investments. One project
can redefine an entire citys trajectory (a no return departure from the baseline) and carry major
residuals (social, economic, environmental); it is very difficult assess ex ante all these effects and
their subsequent impacts on GHGs, accessibility, equity, etc. Standard assessment techniques will
not likely suffice either.
3.6. Final Words on Tools, Techniques, Capabilities
Ultimately, we know that no model of the complex land use-transport system can accurately
predict what will happen over a 25 year period. Models can really only be used to predict the
12
range of possible futures and the impacts that policies today may have on the future; estimates
have been shown to vary significantly depending on the modelling approach used (e.g., Hunt, et
al., 2001). Error compounding in the modeling process can result in errors larger than the
differences between performance of the alternatives (Wachs, 1998). One problem is we dont
really know, in an ex-post sense, how good our predictive capabilities are; ex-post assessments
tend to relate to, for example, an apparent propensity to over-predict public transport ridership or
other demand forecasts and costs (Pickrell, 1992; Skamris & Flyvberg, 1997).
Ex-post assessment of the accuracy of these models is rarely carried out, in part because
of the difficulty in assessing a models effectiveness for predicting a future that almost always
differs from the future that actually plays out (e.g., due to differences in projected vs. actual
economic/demographic growth rates, fuel prices, implementation of different projects or policies
than expected, etc.). In Santiago de Chile, transportation model (ESTRAUS, considered the state
of the art, in practice) runs in 1990-91 were used to evaluate a proposed new Metro Line, the
model over-estimated morning peak loads by 15 to 20% upon line opening in 1997; this overestimation can be at least partially attributed to the fact that the original model runs had been
based on (1) higher relative Metro fares than those actually charged at the time of modeling and
(2) feeder bus services at the southern terminal that were never implemented (de Cea, et al, 2003).
None of this is meant to discredit the need for analysis. We will always be better off with
better analytical tools, and the data-gathering, analysis and understanding, exploration, and
learning and new insights that will almost certainly come with them. And, in many cases,
analytical tools will at least be able to tell us the directionality, if not the exact size of the change
in many of the indicators of interest. But, when we are concerned with precise estimates
particularly of, e.g., GHG emissions for the purpose of financing a project based on a future flow
of carbon revenues then the question of how to account for the inevitable uncertainty arises.
Will the investor force improved ex-ante analysis (screening white elephants)? Can we monitor
year-by-year effects (via, e.g., stratified surveys) in conditions where almost everything else is
also changing? Should the international community cut transport some slack in terms of the
rigor of the required estimates due to its (possibly) crucial role in sustainability? If so, how do we
define and measure the broader sustainability?
4. REFERENCES
Barrett, J. and A. Scott. 2003. The Application of the Ecological Footprint: a case of passenger
transport in Merseyside. Local Environment, Vol. 8, No. 2, pp. 167183.
Bhat, C., S. Handy, K. Kockelman, H. Mahmassani, Q. Chen, L. Weston. 2000. Accessibility
Measures: Formulation Considerations and Current Applications. Research Report No. 7-4938-2.
13
Conducted for the Texas Department of Transportation by the Center for Transportation
Research, University of Texas at Austin, September.
Black, W.R. 2000. Toward a Measure of Transport Sustainability. Paper prepared for
presentation at the 79th Annual Meeting of the Transportation Research Board, Washington, DC,
9-13 January.
Browne, J., E. Sanhueza, E. Silsbe, S. Winkelman, C. Zegras. 2005. Getting on Track: Finding a
Path for Transportation and the CDM. International Institute for Sustainable Development,
Winnipeg, Manitoba.
California Air Resources Board (CARB). 2004. California's Air Quality History Key Events. Last
updated 28 May 2004, last accessed 15 July 2006 at:
http://www.arb.ca.gov/html/brochure/history.htm.
de Cea, J., J.E. Fernndez, V. Dekock, A. Soto. 2003. ESTRAUS: A Computer Package for
Solving Supply-Demand Equilibrium Problems on Multimodal Urban Transportation Networks
with Multiple User Classes. Paper presented at the Annual Meeting of the Transportation
Research Board, Washington, DC.
Daly, H.E. 2002. Sustainable Development: Definitions, Principles, Policies. Invited Address,
World Bank, Washington, DC, April 30.
Dong, X., M. Ben-Akiva, J. Bowman, J. Walker. 2006. Moving from Trip-Based to ActivityBased Measures of Accessibility. Transportation Research A, Vol. 40, no. 2, pp. 163-180.
Geurs, K.T. and B. van Wee. 2004. Accessibility evaluation of land-use and transport strategies:
review and research directions. Journal of Transport Geography, Vol. 12, pp. 127-140.
Goldman, T. & R. Gorham. 2006. Sustainable urban transport: Four innovative directions.
Technology in Society, vol. 28, pp. 261-273.
Hunt, J.D. 2003. Modeling Transportation Policy Impacts on Mobility Benefits and KyotoProtocol-Related Emissions. Built Environment, Vol. 29, No. 1, pp. 48-65.
Hunt, J.D., RA Johnston, JE Abraham, CJ Rodier, G. Garry, SH Putnam, T de la Barra. 2001.
Comparisons from Sacramento Model Test Bed. Transportation Research Record 1780, pp. 5363.
Jeon, C.M and A. Amekudzi. 2005. Addressing Sustainability in Transportation Systems:
Definitions, Indicators, and Metrics. Journal of Infrastructure Systems, March, pp. 31-50.
Kennedy, C.A. 2002. A comparison of the sustainability of public and private
transportation systems: Study of the Greater Toronto Area. Transportation 29, pp. 459493.
Kennedy, C., E. Miller, A. Shalaby, H. Maclean, J. Coleman. 2005. The Four Pillars of
Sustainable Urban Transportation. Transport Reviews, Vol. 25, No. 4, pp. 393-414.
Lautso, K., and Toivanen, S. 1999. The SPARTACUS System for Analyzing Urban
Sustainability. Paper presented at the 79th Annual Meeting of the Transportation Research Board,
Washington, DC, January.
14
15
Serageldin, I. 1996. Sustainability and the wealth of nations: first steps in an ongoing journey.
Environmentally Sustainable Development Studies and Monographs Series No. 5, World Bank,
Washington, D.C.
Skamris, M.K. and B. Flyvbjerg. 1997. Inaccuracy of traffic forecasts and cost estimates on large
transport projects. Transport Policy, 4, No. 3, pp. 141-146.
United Kingdom Commission for Integrated Transport (UK CFIT). 2004. A Review of Transport
Appraisal: Advice from the Commission for Integrated Transport. 4 October, Last accessed 18
May, 2005 at: http://www.cfit.gov.uk/reports/rta/pdf/rta.pdf.
United Nations Division for Sustainable Development (UN DSD). 1992. Agenda 21. UN
Department of Economic and Social Affairs. Last accessed 15 July 2006 at:
http://www.un.org/esa/sustdev/documents/agenda21/english/agenda21toc.htm.
Wachs, M. 1998. The Functions of Models and Analysis in the Policy Process. In Transportation
Models In the Policy-Making Process: Uses, Misuses, and Lessons for the Future, Proceedings
from a symposium on the problems of transportation analysis and modeling in the world of
politics, Asilomar Conference Center Pacific Grove, California, March 4-6, pp. 3-4.
Weaver, R. 1965. Planned Communities. Highway Research Record Number 97, pp. 1-6.
World Bank. 1996. Sustainable Transport: Priorities for Policy Reform. Washington, DC.
World Business Council for Sustainable Development (WBCSD). 2001. Mobility 2001: World
Mobility at the End of the Twentieth Century and its Sustainability. Prepared by the
Massachusetts Institute of Technology and Charles River Associates for the WBCSD Sustainable
Mobility Working Group.
Yevdokimov, Y. 2004. Sustainable Transportation in Canada. Draft paper. Departments of
Economics and Civil Engineering, University of New Brunswick, Canada.
Zegras, C. 2005. Sustainable Urban Mobility: Exploring the Role of the Built Environment.
Unpublished doctoral dissertation, Department of Urban Studies and Planning, Massachusetts
Institute of Technology, September, available at:
http://web.mit.edu/czegras/www/Final%20VersionV3.pdf.
Zegras, C., I. Poduje, W. Foutz, E. Ben-Joseph, O. Figueroa. 2004. Indicators for Sustainable
Urban Development. Chapter 7 in From Understanding to Action: Sustainable Urban
Development in Medium-Sized Cities in Africa and Latin America (M. Keiner, C. Zegras, W.
Schmid, D. Salmern, Eds.), Springer, Dordrecht, the Netherlands.
Zietsman, J. and L.R. Rilett. 2002. Sustainable Transportation: Conceptualization and
Performance Measures. Report 167403, Texas Transportation Institute, The Texas A&M
University System, College Station, Texas, March.
16