National Academies Press: OpenBook

Strategic Issues Facing Transportation, Volume 6: The Effects of Socio-Demographics on Future Travel Demand (2014)

Chapter: Chapter 4 - Study Approach: Scenario Planning and System Dynamics Model

« Previous: Chapter 3 - Key Trends, Drivers, and Projected Impact on Travel Behavior
Page 23
Suggested Citation:"Chapter 4 - Study Approach: Scenario Planning and System Dynamics Model." National Academies of Sciences, Engineering, and Medicine. 2014. Strategic Issues Facing Transportation, Volume 6: The Effects of Socio-Demographics on Future Travel Demand. Washington, DC: The National Academies Press. doi: 10.17226/22321.
×
Page 23
Page 24
Suggested Citation:"Chapter 4 - Study Approach: Scenario Planning and System Dynamics Model." National Academies of Sciences, Engineering, and Medicine. 2014. Strategic Issues Facing Transportation, Volume 6: The Effects of Socio-Demographics on Future Travel Demand. Washington, DC: The National Academies Press. doi: 10.17226/22321.
×
Page 24
Page 25
Suggested Citation:"Chapter 4 - Study Approach: Scenario Planning and System Dynamics Model." National Academies of Sciences, Engineering, and Medicine. 2014. Strategic Issues Facing Transportation, Volume 6: The Effects of Socio-Demographics on Future Travel Demand. Washington, DC: The National Academies Press. doi: 10.17226/22321.
×
Page 25
Page 26
Suggested Citation:"Chapter 4 - Study Approach: Scenario Planning and System Dynamics Model." National Academies of Sciences, Engineering, and Medicine. 2014. Strategic Issues Facing Transportation, Volume 6: The Effects of Socio-Demographics on Future Travel Demand. Washington, DC: The National Academies Press. doi: 10.17226/22321.
×
Page 26
Page 27
Suggested Citation:"Chapter 4 - Study Approach: Scenario Planning and System Dynamics Model." National Academies of Sciences, Engineering, and Medicine. 2014. Strategic Issues Facing Transportation, Volume 6: The Effects of Socio-Demographics on Future Travel Demand. Washington, DC: The National Academies Press. doi: 10.17226/22321.
×
Page 27
Page 28
Suggested Citation:"Chapter 4 - Study Approach: Scenario Planning and System Dynamics Model." National Academies of Sciences, Engineering, and Medicine. 2014. Strategic Issues Facing Transportation, Volume 6: The Effects of Socio-Demographics on Future Travel Demand. Washington, DC: The National Academies Press. doi: 10.17226/22321.
×
Page 28
Page 29
Suggested Citation:"Chapter 4 - Study Approach: Scenario Planning and System Dynamics Model." National Academies of Sciences, Engineering, and Medicine. 2014. Strategic Issues Facing Transportation, Volume 6: The Effects of Socio-Demographics on Future Travel Demand. Washington, DC: The National Academies Press. doi: 10.17226/22321.
×
Page 29

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

23 The goal of this research is to provide transportation planners and decision makers with an increased awareness of socio-demographic trends and how these may impact long-range transportation conditions or needs. This capability is delivered through the written findings and principles compiled in this final report, and also through analysis based on the application of a tool comprised of a SD model and scenarios. This joint scenario/modeling approach is unique. The SD model simulates the demographic evolution of a regional population through 2050, while indicating the impact on travel demand and considering the influence of employment, land-use, and transport supply sectors. The scenarios are exogenous to the SD model. Their influ- ences on demographic evolution and travel demand can be explored by manipulating SD model parameters that are linked to scenario assump- tions. Together, the scenarios and the SD model provide an understand- ing of the fundamental relationships between social and demographic factors and travel demand, and how these relationships may change over time. The model more fully develops the socio-demographic sector and linkages to travel demand better than other models in the area of model- ing. Detailed documentation on the validity of the internal relationships in the model can be found in Appendix B. 4.1 Rationale for Approach The rationale for the joint scenario/modeling approach developed from the team’s review of methods used to forecast travel demand, focusing in particular on how those methods account for changes in socio-demographic factors. Travel demand forecasting is a broad topic, and the team has not attempted to cover all aspects in this document. For readers interested in the state of the practice in travel demand forecasting, we recommend Donnelly et al. (2010). The team considered five general types of models in its review: (1) project-specific models, (2) regional models, (3) statewide models, (4) strategic models, and (5) land-use models. A brief description of each is given below. • Project-specific models—First developed in the transportation field to predict the effects of road capacity projects on traffic flows and congestion, project-specific models remain the type of model most commonly used today. They typically focus on a specific transportation C H A P T E R 4 Study Approach: Scenario Planning and System Dynamics Model Chapter 4 Takeaways • This research applies a unique joint scenario/modeling approach. • Scenario planning is an effective way for agencies to deal with complexity and uncertainty. • System dynamics methodology is well suited to the needs of various analytical challenges in transportation. “The only relevant discussions about the future are those where we succeed in shifting the question from whether something will happen to what would we do if it happened.” Arie de Geus, former coordinator, Group Planning, Shell International Petroleum Company

24 The Effects of Socio-Demographics on Future Travel Demand corridor and set of intersections, with the assumption that the changes being modeled will not significantly affect wider regional traffic patterns. In most cases, the number of trips entering the area at any given time of day is taken as a given, based on observed counts, so the demo- graphic profile of travelers and possible changes in travel behavior (shifts in travel mode, destination, trip frequency, etc.) are not modeled in detail. • Regional models—Every U.S. MPO is required to prepare a regional transportation plan (RTP) periodically to qualify for federal (and state) transportation funds and to demonstrate compliance with environmental regulations. The time horizon of these forecasts is typically 25 years, and the MPO region may cover a single county or several counties. Regional model- ing is where a great deal of methodological development is taking place. In recent years, the shift has been away from aggregate, zone-based models with a very limited amount of demo- graphic detail toward disaggregate, parcel-based models that simulate the travel behavior of each individual in a detailed, representative population database. Currently, most regional travel demand models are used for forecasting, and, in most cases, little attention is paid to the level of uncertainty in the forecasts and the possible risk entailed. • Statewide models—State DOTs are responsible for providing and maintaining major high- ways and rural roads, which are the main intercity travel connections (and sometimes the major intracity corridors as well). Most states maintain a statewide travel demand forecasting model to assist in planning highway capacity. Statewide models tend to be more aggregate and less behaviorally detailed than regional models. But like the regional models, they do not explicitly address uncertainty. • Strategic models—Strategic models are used to consider large policy or imposed changes on the population (e.g., fuel prices). They are sometimes developed within an SD framework, sometimes as a stand-alone program, usually without explicit transport network loading, but sometimes with network supply effects being modeled. This is an emerging trend in long- range planning, where there is awareness that one cannot actually forecast the future, and many scenario possibilities need to be studied so that a policy or investment strategy that minimizes risk, or moves toward some desired goal(s), can be followed. The need to consider many scenarios implies “fast” models. These models also tend to take into account the path through time into the future (path-based), rather than the traditional “end-state” approach followed in most local, regional, and statewide travel forecasting models. • Land-use models—Land-use modeling can be thought of as a complement to travel demand modeling—a very important complement with respect to demographics. Land-use models tend to be run dynamically through time, keeping track of the types of buildings and residential and commercial uses of each parcel of land, and simulating the transactions of the property market as subject to zoning restrictions and economic forces. Ideally, a good land-use model will be integrated with a good travel demand model, with the land-use model predicting how differ- ent types of people, households, businesses, and jobs will be distributed throughout the region, and the travel demand model predicting the travel patterns and traffic congestion that result. Table 4-1 provides a side-by-side comparison of the typical characteristics of these models. The research team’s review concluded that none of the travel demand models currently applied in the United States uses a fully detailed demographic evolution model. But, the strategic or integrated land-use models seemed to be most relevant for this study’s purposes. In addition, the team surmised that it was important to consider the dynamic feedback loops between travel demand and socio-demographic factors. Not only do demographic and socioeconomic factors influence travel behavior, but travel behavior (as it manifests in the aggregate) can, in turn, influ- ence a region’s socioeconomic and demographic profiles. The review of travel demand modeling and forecasting was extended to examine emerg- ing issues and evolving models, and to address accuracy and uncertainty in travel forecasts.

Study Approach: Scenario Planning and System Dynamics Model 25 This examination deduced that advanced, activity-based travel demand models have adopted an agent-based, micro-simulation approach, with greater spatial and temporal detail to better handle emerging issues. This approach is more realistic than older, aggregate-based approaches, and should lead to increased accuracy for certain short-term types of modeling and prediction. However, it was concluded that the agent-based, micro-simulation approaches will not lead to more accurate predictions when considering a range of future possibilities, which is the context for the research. Thus, the team arrived at the decision to adopt an SD modeling framework that illustrates feedback loops, which is complemented by a scenario approach to account for uncertainties in forecasts. Before we explain our SD modeling approach, however, it is beneficial to discuss our rationale for using scenario planning. 4.2 Scenario Planning Approach Chapter 3 identified key socio-demographic trends that will impact travel demand over the next 30–50 years. These trends are based primarily on straightlining of current trends carried into the future and also on analysis and projections based on known factors. Transportation agencies face strategic decisions in terms of how to cope with or adapt to these trends and their impact on travel demand. These are complex questions that depend on a variety of difficult-to- predict factors beyond the control of transportation agencies. Also, the impacts of the trends on future travel demand are to a large extent uncertain, incomplete, evolving, or conflicting. Model Characteristics Model Types Considered Project-Specific Regional Statewide Strategic Land-Use Typical areas of focus Road design, traffic flow, level of service. Policy costs and benefits, air quality. New Infrastructure (e.g., transit or a new road) Main highways, freight, longer- distance travel. Broader range of strategies. Residential and commercial land, zoning. Typical time horizon 0–10 years 10–25 years 10–30 years 20–50 years 5–30 years Typical spatial boundary A corridor or neighborhood. A region or county. A state or group of states. Varies. A region or county. Typical level of spatial detail Fine level. All road links and intersections. Zones are blocks or larger. Only major roads, arterials. Broader detail. Only major intercity roads. Broader detail. Spatial abstraction. Individual parcels or grid cells. Travel demand relationships included Few. Use fixed demand. Focus on traffic flow. Advanced models include a broad range of behavior. Some, but usually less detail than regional model. Focus on key behavioral aspects of interest. Few. Focus is on land market behavior. Demographic detail typically included Very little. Advanced models include a wider range of variables. Some, but usually less detail than regional model. As much as needed for the topic area. Can include a wide range of variables. Table 4-1. Overview of travel demand model types considered.

26 The Effects of Socio-Demographics on Future Travel Demand Scenario planning is one way transportation agencies can deal with such complexities. Sce- narios are generally a way of thinking about the future. Scenario planning is typically defined as a process of surfacing a set of plausible alternative futures, determining a range of possible consequences, and identifying strategies or policy options that would be robust across the set of futures (Lempert et al. 2003). Most authors attribute the introduction of scenarios to Herman Kahn through his work for the U.S. military in the 1950s at the RAND Corporation, where he developed a technique of describing the future in stories as if written by people in the future. He adopted the term “scenarios,” which was originally used in the context of performing arts, to describe these stories (Chermack et al. 2001 and Khan 1965). Today there is a rich variety of scenario approaches, reflecting different aims and interests and the characteristics of different fields of application. Among others, two types of applications have brought the scenario technique to the forefront in recent years. On the one hand, there has been the production of global scenarios, whether issue-based, mainly explorative scenarios focusing on climate change, water, etc., or integrated normative visions of the future; and on the other hand, there have been more locally scaled scenarios focusing on the potential of development of a specific region or city. In addition, varying uses or ways of thinking about the future are often interrelated with the term scenario (Sparrow 2000, Dreberg 2004): • Sensitivity analysis is a technique typically used to determine how different values of an inde- pendent variable will affect a particular dependent variable under a given set of assumptions, and is often used in mitigating risks in forecasts. But in effect, sensitivity analysis is more akin to a predictive mode of thinking about the future than uses below. • Contingency planning is a plan devised for an outcome other than in the usual (expected) plan, and is often used for military or civil emergency planning. But contingency planning could also be applied to decision making in corporate or public policy. • Normative approach seeks to envisage how society or some sector or activity could be designed in a better way than its present mode of functioning. This approach identifies solutions to fundamental societal problems by defining normative goals and exploring the paths leading to these goals. • Exploratory analysis is in the form of coherently structured speculation, so many different developments or possible events can be described. Its strategic purpose is to better prepare decision makers and planners to handle emerging situations, recognizing that it is impossible to predict what will actually happen. Many methodologies are available to develop scenarios, as discussed in Amer et al. (2013). There is no single “correct” method, and different contexts require different scenario methods. The scenario methodology used in this project was based on expert elicitation and strategic assumption surfacing, as noted in detail in Chapter 5. The team started the scenario development process by identifying the key factors (i.e., influ- encers) on travel demand, in order to develop a scenario framework. Such a framework pro- vides a clear logic and structure for describing the scenarios and differentiating them from one another. The initial catalog of socio-demographic drivers was organized into six categories: (1) population size and growth, (2) geo-demographics of population size and growth, (3) pop- ulation structure and composition, (4) household-based economic activity, (5) cultural and social diversity, (6) and external factors intertwined with socio-demographics—i.e., urban form, technology, infrastructure investment. These drivers formed the basis of the scenario framework (see Table 5-1 in Chapter 5). Using this framework as an organizing structure, the team produced four scenarios of hypothetical futures that were distinct from each other and yet internally consistent, as discussed in Chapter 5.

Study Approach: Scenario Planning and System Dynamics Model 27 Using the framework to develop the scenarios ensured that the scenarios focused on the joint effect of many factors and, thus, captured both the complexity and the uncertainties inherent in describing the impact of socio-demographics on travel demand. Relating this to the trends presented in Chapter 3, scenarios can help decision makers and planners to understand how the various strands of the complex tapestry of socio-demographic trends will move if one or more threads (or trends) are pulled in one direction versus another. When the various factors are con- sidered together, one realizes that certain combinations could magnify each other’s impact or likelihood. For instance, an increased trade deficit may trigger an economic recession, which in turn creates unemployment among young adults who put off forming new households by moving in with their parents, which then reduces VMT in a region. Despite the growing popularity of scenario planning, a number of misconceptions remain about what it is, when it should be used, how it should be deployed, and what benefits it can yield. All too often, scenario approaches deteriorate into little more than a conventional fore- casting effort that involves assigning explicit probabilities to potential outcomes. Or, at the other extreme, scenario planning devolves to loosely grounded futurist musings with little if any relevance to current circumstances. The art of scenario-based, long-term planning is to connect the work of “what ifs” with down- to-earth decision-making processes. The proposed joint scenario/modeling approach is designed to do just that. The SD model enables the running of many different scenarios to examine travel behavior outcomes at different points in time. So Impacts 2050 can integrate the scenarios into the long-range planning process of DOTs and MPOs. 4.3 SD Modeling Approach While the objectives of the study did not explicitly call for a “model” as a product, the research team believed that the stated goal could not have been achieved without creating and using such an analytical tool. To recap, in arriving at the modeling approach the team recognized the following: • Conventional transportation planning models used for estimating travel demand tend either to be limited in scope in terms of the types of variables that are represented, or to have very long run times (hours, days, or even weeks), which limits the range of assumptions that can be practically tested for long-range forecasting. • It is unrealistic to believe that any traditional forecasting procedure could predict future out- comes with a high degree of certainty, given the large uncertainty in the inputs, the many factors at play, and the inter-relationships that may change over time. For example, a 30-year simulation of the land-use and activity-based travel model of the San Francisco Bay area would require 216 hours, or nine days of computation time (Waddell et al. 2010). • Given these issues, the team decided to use an SD modeling approach because it not only is a method designed for scenario testing, leading to understanding how and why things may change over time, but is also an approach that facilitates rapid, “hands-on” analysis of many scenarios. SD was first developed by Jay Forrester at Massachusetts Institute of Technology in the 1960s. One of his first applications was the longer-term evolution of urban land use and population, as described in his book Urban Dynamics (Forrester 1972). The SD approach has developed and evolved since then, and has been applied in dozens of different fields, includ- ing business management, economics, biology, the physical sciences, and various aspects of the social sciences. These models are typically developed using recent time series data for key

28 The Effects of Socio-Demographics on Future Travel Demand variables, and calibrating the model to reproduce the time trends for those variables reason- ably well. Because, as described below, the focus of such models is typically on exploring scenarios and policies, rather than producing long-term forecasts, there has been very little emphasis in the literature on exploring the predictive accuracy of the models over time. One exception is the model used for the well-known “limits to growth” study in the 1970s. Updates to that study have been published to show how well the model has been able to track actual trends over time (Bardi 2011). Although still relatively new in transportation modeling, the SD approach has been applied to address similar objectives as in this study. For example, Fiorello et al. (2006) developed an SD model that covers the European Union and larger Europe. The model includes a population module that evolves the population for 29 countries by age in one-year increments through cohort survival, in-migration, and intermigration. In the United States, an SD model was devel- oped to evaluate policies for development of the Las Vegas Valley (Stave and Dwyer 2006). This model deals specifically with the population by age, including births, deaths (cohort survival), and migration based on “attractiveness” for a large number of alternative scenarios in a path- based analysis. Also, there have been a few cases over the years where the urban dynamics model has been updated and extended to deal with transportation and land-use interactions. But in general, the SD approach has not been used extensively in modeling travel behavior (a range of applications can be observed by looking at the list of conference proceedings for the System Dynamics Society at www.systemdynamics.org/society_activities.htm). Travel demand modeling has tended to develop its own network-based, static equilibrium- based modeling approaches. Generally, these methods have been developed independently from approaches used in other fields. While travel demand modeling methods have their strengths— particularly the more recent activity-based approaches—the SD approach seemed more appro- priate to use for this project for a number of reasons. Two main aspects relate to model fidelity and are termed structural accuracy and statistical accuracy: • Structural accuracy refers to having the correct set of variables in the model, and the correct set of causal relationships linking those variables. • Statistical accuracy refers to the exact numerical parameters and functional forms used in defining the model relationships. Generally, the emphasis in travel demand modeling has been on statistical accuracy, using a fairly limited set of variables and relationships, with most of the effort devoted to estimating and calibrating the parameters of the models. Such an emphasis may be appropriate for shorter- term forecasts, where many variables and relationships can be assumed to remain constant or to be exogenous and one-directional—i.e., they influence travel behavior, but changes in travel behavior do not feed back to influence them. The farther that one expands the time horizon and the range of scenarios under consideration, the more important structural accuracy becomes relative to statistical accuracy. When one is predicting far out into the future, the range of uncertainty in the input variables becomes larger and larger, and the use of precise statistical relationships becomes less and less relevant. The emphasis becomes less on numerically accurate forecasts and more on qualitatively accurate depictions of how different variables and relationships will evolve over time. Correspondingly, the approach for calibrating and validating a model depends less on obtaining exact matches to a limited number of data items and more on trying to match the qualitative trends that have been observed in the real world over a length of time (e.g., running a model simulation from 1970 to 2010 and comparing the “predicted” trends with actual trends over time).

Study Approach: Scenario Planning and System Dynamics Model 29 The SD modeling approach is focused on model structure, using a framework based on physi- cal entities that build up or diminish over time (stocks), the rates of change in those entities (flows), and the relative timing of those changes (delays). The focus is also on feedbacks within the system—whether various modeled relationships tend to reinforce each other (“positive” feedback), or work in opposition to each other (“negative” feedback). For longer-term models, this distinction is critical, because negative feedback relationships tend to lead to constrained behavior that is fairly stable over time, while positive feedback relationships, although rarer, can lead to exponential growth or other forms of unstable behavior over time. As a result, the model structure makes it so that the predicted behavior is more sensitive to some variables and relationships than others, even if one would not expect so from looking at the relative size of the numbers used to parameterize the relationships. Describing a complex system in this way is often useful in illustrating how seemingly simple rules may result in a complex, nonlinear system (Pfaffenbichler 2011). The critical importance of the dynamic model structure, even more than the model param- eters, may seem like a strange concept, particularly to those who have worked in the static world of travel demand modeling for some time. However, for a model with a long time horizon and a wide scope, this way of thinking can be very valuable. The main purpose of the model used in this study is not to provide long-term forecasts— without a crystal ball, those forecasts would almost certainly be wrong and not very useful. Rather, the model’s main purpose is to facilitate the running of many different scenarios. SD models typically do not model transport network loading explicitly, but include some simple representations of network supply effects. This approach is proven to greatly reduce model run times—typical SD models are capable of producing 50-year forecasts in less than a minute— which makes the exploration of a large number of scenario tests possible. There are many examples of the value of running many different scenarios. For example, transportation agencies can visualize how and why various possible futures may occur and reflect on how political, social, and economic changes may affect operations and plan accord- ingly. Another example is the fact that when a number of different scenarios are produced, many perspectives can be included and a policy or planning discussion does not have to revolve around the advocacy of fixed positions. Finally, in running many different scenarios, issues may be sur- faced by exposing the underlying forces in a region that otherwise would not be considered in the planning process. The bottom line is that the analysis of these scenarios can assist in decision making and resource allocation. Chapter 5 provides more detailed information about the scenarios developed, and Chapter 6 presents Impacts 2050, the underlying SD model structure, and its outputs.

Next: Chapter 5 - Future Scenarios: Process and Narratives »
Strategic Issues Facing Transportation, Volume 6: The Effects of Socio-Demographics on Future Travel Demand Get This Book
×
MyNAP members save 10% online.
Login or Register to save!
Download Free PDF

TRB’s National Cooperative Highway Research Program (NCHRP) Report 750: Strategic Issues Facing Transportation, Volume 6: The Effects of Socio-Demographics on Future Travel Demand presents the results of research on how socio-demographic changes over the next 30 to 50 years may impact travel demand at the regional level. It is accompanied by a software tool, Impacts 2050, designed to support the long-term planning activities of transportation agencies.

The print version of the report contains a CD-ROM that includes Impacts 2050, the software user’s guide, a PowerPoint presentation about the research, and the research brief. The CD-ROM is also available for download from TRB’s website as an ISO image. Links to the ISO image and instructions for burning a CD-ROM from an ISO image are provided below. This is a large file and may take some time to download using a high-speed connection.

Help on Burning an .ISO CD-ROM Image

Download the .ISO CD-ROM Image*

NCHRP Report 750, Volume 6 is part of a series of reports being produced by NCHRP Project 20-83: Long-Range Strategic Issues Facing the Transportation Industry. Major trends affecting the future of the United States and the world will dramatically reshape transportation priorities and needs. The American Association of State Highway and Transportation Officials (AASHTO) established the NCHRP Project 20-83 research series to examine global and domestic long-range strategic issues and their implications for state departments of transportation (DOTs); AASHTO's aim for the research series is to help prepare the DOTs for the challenges and benefits created by these trends.

Other volumes in this series currently available include:

• NCHRP Report 750: Strategic Issues Facing Transportation, Volume 1: Scenario Planning for Freight Transportation Infrastructure Investment

• NCHRP Report 750: Strategic Issues Facing Transportation, Volume 2: Climate Change, Extreme Weather Events, and the Highway System: Practitioner’s Guide and Research Report

• NCHRP Report 750: Strategic Issues Facing Transportation, Volume 3: Expediting Future Technologies for Enhancing Transportation System Performance

• NCHRP Report 750: Strategic Issues Facing Transportation, Volume 4: Sustainability as an Organizing Principle for Transportation Agencies

• NCHRP Report 750: Strategic Issues Facing Transportation, Volume 5: Preparing State Transportation Agencies for an Uncertain Energy Future

*CD-ROM Disclaimer - This software is offered as is, without warranty or promise of support of any kind either expressed or implied. Under no circumstance will the National Academy of Sciences or the Transportation Research Board (collectively "TRB") be liable for any loss or damage caused by the installation or operation of this product. TRB makes no representation or warranty of any kind, expressed or implied, in fact or in law, including without limitation, the warranty of merchantability or the warranty of fitness for a particular purpose, and shall not in any case be liable for any consequential or special damages.

  1. ×

    Welcome to OpenBook!

    You're looking at OpenBook, NAP.edu's online reading room since 1999. Based on feedback from you, our users, we've made some improvements that make it easier than ever to read thousands of publications on our website.

    Do you want to take a quick tour of the OpenBook's features?

    No Thanks Take a Tour »
  2. ×

    Show this book's table of contents, where you can jump to any chapter by name.

    « Back Next »
  3. ×

    ...or use these buttons to go back to the previous chapter or skip to the next one.

    « Back Next »
  4. ×

    Jump up to the previous page or down to the next one. Also, you can type in a page number and press Enter to go directly to that page in the book.

    « Back Next »
  5. ×

    To search the entire text of this book, type in your search term here and press Enter.

    « Back Next »
  6. ×

    Share a link to this book page on your preferred social network or via email.

    « Back Next »
  7. ×

    View our suggested citation for this chapter.

    « Back Next »
  8. ×

    Ready to take your reading offline? Click here to buy this book in print or download it as a free PDF, if available.

    « Back Next »
Stay Connected!