California Department of Transportation

Adobe Acrobat Reader is required to view and print PDF files. Adobe Acrobat is required to edit online forms and save them to your computer.


Last Updated: Thursday, March 24, 2016 4:07 PM

Analytical Tools

The following discussions are intended to help the practitioner assess which analytical tool or combination of tools may be appropriate to use when analyzing the growth-related effects of a highway project. Several tools are described – qualitative analysis; transportation forecasts; geographic information systems (GIS); integrated land use and transportation models; and regression analysis, econometric forecasting techniques, and models. The discussions include the basic types of each tool, when they might be applied, their strengths and weaknesses, and sources for additional information.

Qualitative Analysis

Qualitative methods using expert knowledge are used frequently to predict and evaluate land use interactions. There are a variety of qualitative analysis methods that can be applied to growth-related impact analyses. In general, qualitative approaches are most effective if used in conjunction with quantitative and GIS-based methods. Similarly, quantitative methods nearly always require the framework and context that an effective qualitative study provides. Gathering expert opinions and qualitative analysis can be helpful in developing a more focused analysis of known issues and can help frame corresponding quantitative and/or GIS studies.

Basic Types

There are several broad categories of qualitative techniques:

  • Stakeholder and Focus Group Meetings – This approach uses engagement with locally affected citizens and experts to gather background information, knowledge of key issues and to find what resources are considered most valuable to neighborhoods affected by a given project.
  • Qualitative Inference – This technique involves a case study description of an area of concern, e.g., habitat or neighborhood, and an identification of resources based on professional judgment of the possible impacts that the proposed project would entail. The case study focuses on the indicators that characterize the area of concern. Techniques involving professional judgment are often combined with other techniques noted here.
  • Literature Review/Comparative Case Analysis – A comparative study involves comparing a like area where a similar project has been completed to the area of concern where a project is proposed. This is similar to the Qualitative Inference approach, but uses comparisons to enrich the analysis.
  • The Delphi Method – This is a more systematic way to use expert opinions based on an interviewing method that begins with general questions, but focuses the questions and the analysis more precisely as the process continues. It employs survey research technique directed toward the systematic solicitation and organization of expert intuitive thinking from a group of knowledgeable people. It entails elements of the two methods described above, but is a more structured process.
  • Scenario Writing – This method creates an outline in narrative form of some conceivable future environment given certain assumptions about the present and a sequence of events in the intervening period. Multiple scenarios can be drafted to include a variety of changing conditions, a spectrum of potential developments, and a series of hypothetical socio-political, ecological, and economic consequences of proposed actions. This technique can develop ideas and identify causal relationships that might not surface in more structured methods. Rather than predictive, scenario writing is a technique which attempts to establish some logical sequence of events to show how, under present conditions and assumptions, a future environment might evolve. Scenarios can also serve to set the upper and lower bounds of potential outcomes.
  • Networks – Creating system diagrams or networks can be used in classifying, organizing, and displaying problems, processes, and interactions and to produce a causal analysis of the indirect/cumulative effects. This approach is a diagrammatic version of the scenario writing method and assumes a high level of knowledge and expertise by its designer. The Network approach can be both qualitative and quantitative.
  • Matrices – This technique can assist in the display and interpretation of information developed using many other qualitative and quantitative techniques. The matrix is commonly a grid diagram in which two distinct lists are arranged along perpendicular axes, e.g., actions and environmental characteristics. The interactions between the two that would produce impacts are noted and effects are described in a binary fashion (yes or no) a qualitative fashion (descriptive paragraph) or a quantitative fashion rank or index.

Typical Applications

Qualitative methods can usefully serve to evaluate the context or overall situation wherever little historical data exist or wherever existing data are questionable or inconsistent. In most cases, qualitative approaches to an impacts assessment are part of a larger, multi-pronged approach to doing an analysis. Qualitative approaches are most important for their ability to help frame an impact analysis. This is a most critical function when designing very large and complex analyses.

Strengths and Weaknesses

Strictly qualitative approaches have some limitations and risks. Foremost among these is the risk of slipping into speculation based on limited data or unusual circumstances. Broad participation, including input from local planners, experts, or other stakeholders through surveys, interviews, or task forces can serve as a safeguard against this. Broad and diverse participation also serves to protect against ideological biases, which is a risk when relying heavily on qualitative analyses. The Scenario Writing and Network methods are only as good as the underlying understanding or assumptions of often complex processes and interactions. Similarly, they place a high bar on the knowledge and expertise of the practitioners crafting them.

This summary was adapted from NCHRP Report 466, Course Modules 7 and 8.

For More Information

Linstone, H.A. and M. Turoff, eds. 1975. The Delphi Method: Techniques and Applications. Addison-Wesley, Reading MA.

Seskin, S., K. Still, and J. Boroski. April 2002. The Use of Expert Panels in Analyzing Transportation and Land Use Alternatives. NCHRP Projection 8-36 Task 4. Transportation Research Board.

Transportation Forecasts

Transportation planners have long relied on computer-based models to predict how traffic patterns change with improvements to the transportation system. The traditional four-step model estimates how land use results in trips, what type of trips are generated, what mode is used for trips, and where and when those trips occur on the transportation network. Outputs from the travel model can be used to determine to key factors in land use change: accessibility (ease of travel to key destinations) and number of trips (reflecting opportunities for highway oriented or retail businesses).

Basic Types and Typical Applications

There are two basic types of techniques using travel demand models:

  • Screening and Comparative Evaluations – Outputs of a travel demand model (volumes, level of service estimates, travel times, vehicle miles traveled (VMT) can be used to establish the where a project will have an effect on local traffic and travel times and whether the effect is regional or localized in nature. This involves a forecast of travel demand with a project alternative (build alternative) and without (no-build alternative) and comparison between the two conditions. If a project has a negligible effect on regional travel times, or indicators such as VMT its effect can be determined to be localized. Localized effects can be evaluated by analyzing changes in local traffic conditions in combination with qualitative (e.g., comparative case, scenario writing) or quantitative (e.g., relating traffic levels at a new interchange to types of business that may be supported by that pass-by traffic).
  • Input to Simple Regional Land Use Evaluations – Outputs of the travel model can also be used as input variables to an accessibility analysis (evaluation of how travel times between key destinations change with and without a project) and a simple gravity model (method for allocating growth in households or employment based on accessibility change) for use in a broader regional analysis.· Input to Simple Regional Land Use Evaluations – Outputs of the travel model can also be used as input variables to an accessibility analysis (evaluation of how travel times between key destinations change with and without a project) and a simple gravity model (method for allocating growth in households or employment based on accessibility change) for use in a broader regional analysis.

Strengths and Weaknesses

Transportation Forecasts can provide valuable insight into how a project would affect local and regional patterns of traffic. Analysts can use this information in qualitative or quantitative assessments to establish the location and extent to which changes in accessibility may affect land use change. Traditional models will not provide direct output of the key variables (households and employment) in an indirect impact evaluation and will not capture the dynamic interaction of land use and transportation in feedback loops over time (see the summary for Integrated Land Use and Transportation Models). Travel demand models require special expertise to produce and evaluate results. The expense and complexity of travel models make them appropriate only in situations where an established, calibrated regional or statewide model is in place.

This summary was adapted from NCHRP Report 466, Course Module 8 and NCHRP Report 423A, Section 2, Analytical Tools

For More Information

NCHRP Report 456, Section 2, Changes in Travel Time and Section 6, Accessibility.

Geographic Information Systems (GIS)

GIS provides the ability to map, display, and, analyze spatial data for evaluations of indirect and cumulative impacts. Although cartographic techniques for evaluating impacts have been in use for many years, GIS allows for the assembly of large databases and automated processing. In most locations, state, regional, and local planning agencies maintain GIS datasets that are useful in indirect impact evaluations. These datasets include roadway networks, political boundaries, topography, vacant land, existing land use, zoning, demographic and employment statistics, historic resources, habitats, and natural resources. GIS is useful for all steps in an evaluation but is often combined with other methods.

Basic Types

There are two basic types of techniques using GIS:

  • Map Overlays - The McHarg overlay technique (1969), which involves the combination of project design maps and natural and community feature and resource maps, is time-tested and can be readily implemented in GIS. This technique can be particularly useful for visualizing potential indirect/cumulative effects related to alteration of the physical environment, e.g., habitat fragmentation or community segmentation. GIS has greatly enhanced the ability to process and display cartographic information. Cartographic techniques are limited in their ability to reveal the structure, function, and dynamics of areas. However, their utility can be expanded by relating inventoried information about these characteristics via a relational database.
  • Resource Capability Analysis – Another cartographic technique applicable to identification of indirect/cumulative effects is resource capability analysis (Rubenstein 1987). Similar to the overlay technique, this process involves the preparation of two maps an opportunity map depicting conditions favorable to development (topography, soil types, zoning, and regulatory standards) and a constraint map depicting areas unsuitable for development (wetlands, floodplains, slopes, parklands, or other notable features). Overlaying the two maps produces a land suitability map indicating areas with capacity for potential induced growth. This map could be further modified to indicate areas with the highest potential for complementary development (interchange quadrants) and development shifts (interchanges and feeder roads) under the action alternatives.

Typical Applications

In analyses of growth-related effects, GIS is most often used to catalog resources and identify areas of conflict between features of the project and features of the natural or human environment. These include direct impacts such as property takings or habitat encroachment and indirect impacts to habitats and communities by allowing analysts to determine the location and extant of natural systems and notable community features.

While GIS cannot predict the location of future households or employment, it can be used to determine likely locations for these activities by analyzing the location of existing development, project features, zoning, and natural features and constraints. Some tools are now available which combined GIS input and display capabilities with decision rules or land use modeling techniques to add predictive or scenario evaluation capabilities (see the summary for Integrated Land Use and Transportation Models).

GIS is also particularly effective in displaying the results of evaluations and support data with thematic maps depicting demographics, land use, and socio-economic conditions.

Strengths and Weaknesses

GIS is a widely used, efficient, and effective method for gathering and cataloging data, analyzing spatial data, and documenting assumptions and presenting results. GIS by itself can not be used to develop estimations of land use change and impacts and can not capture the dynamics of many natural and social systems. GIS can be used to support and implement each step in the evaluation process.

This summary was adapted from NCHRP Report 466, Course Module 7.

For More Information

Brail, R.K., and R.E. Klosterman, eds. 2001. Planning Support Systems: Integrating GIS, Models, and Visualization Tools. ESRI Press, Redlands CA.

McHarg, I.L. 1969. Design with Nature. Natural History Press, Washington, D.C.

Rubenstein, H.M. 1987. A Guide to Site and Environmental Planning. John Wiley and Sons, New York, NY.

ESRI (developer of ArcMap GIS software) summary of planning applications located at

Integrated Land Use and Transportation Models

Integrated land use and transportation models represent enhancements to the typical four-step travel demand model used at state and regional agencies. In the traditional models, demographic and land use assumptions used in the estimation of trips are developed outside the model and remain fixed for each forecast year in a model run. Integrated models allow land uses to shift in each forecast year to capture how changes in the transportation system affect land use change, and how land use change will affect volumes on the roadway network. Through an iterative process these integrated models predict an equilibrium land use and traffic pattern for a future year or years in the traffic forecast. Based on region-wide forecasts of population and employment, these models allocate new housing and employment to local areas based on transportation accessibility, land availability, and in some cases land prices and other factors.

Basic Types

There are several basic types of integrated models that vary in their complexity and methods:

  • Scenario Based Models - These models allow the user to enter information about current land use conditions and the transportation network through Geographic Information System (GIS) maps. Users then input parameters on future land use regulations, and weights for factors that typically influence land use decisions. The models then rate land areas for their suitability for development and allocates regional growth to local areas based on suitability. Factor weights and other parameters can then be altered to create scenarios to be compared for planning or impact analysis purposes. Examples include the commercially available What If? and Smart Growth Index packages.
  • Spatial Interaction/Gravity Models – These models use Lowry gravity-model formulation to allocate employment and households based on measures of attractiveness for development including availability of land, travel time and cost, and household income. These models can typically be linked to a region’s travel modeling system to provide a feedback loop. Parameters for allocation are typically estimated through a process of calibration specific to the location being evaluated. Examples include the widely used DRAM-EMPAL/ITLUP/Metropilus package developed by Steven Putman for the U.S. Department of Transportation and the ULAM system used in Florida.
  • Market Equilibrium Models – Several modeling systems in use in the United States and Europe base predictions for household and employment location on the demand and supply for these land uses and information on the economic factors in location choices of households and employers developed through discrete choice estimation techniques. Integration with travel demand models allow the land use models to account for increases or decreases in travel time and cost in location decisions. Parameters for allocation are estimated through a process of calibration specific to the location being evaluated. Examples include UrbanSim, Metrosim, TRANUS, and MEPLAN.
  • Cellular Models – A more recent line of modeling involves making predictions about future land use based on probability modeling developed through time-series observations and decision-rules. One example is LEAM which uses historical series of satellite or aerial photography imagery in combination with maps of attributes and constraints to make predictions on future land uses.

Typical Applications

Most integrated, transportation and land use models require a significant investment in time and money to develop. Most current applications are by academics and Metropolitan Planning Organizations. In areas where these models are already in place and calibrated to local conditions, they can be used to assess the magnitude and location of land use change associated with a transportation improvement. By comparing results using a “no-build” transportation network and a “build” alternative, the analyst can identify the increment of change in households and employment in each area the model analyzes [usually Traffic Analysis Zones (TAZs) made up of census tracts or block groups]. The analyst can evaluate the land use changes in the context of resources and notable features.

Strengths and Weaknesses

Integrated models are based on established theories of location choice and land development. By providing a feedback loop between land use and travel estimation, the models more closely represent reality than tradition travel demand models because they assume a dynamic rather than a static land use/transportation system. Integrated models also allow the analyst to directly estimate the key variables in an induced growth analysis – housing and population. Models available to date, however, have been costly to set-up, implement, and maintain because of their cost, data requirements, and need for calibration to local conditions. For these reasons, these models are used almost exclusively in academic and regional planning settings and there are very limited examples of their use in NEPA/CEQA evaluations of projects.

This summary was adapted from NCHRP Report 466, Course Module 8.

For More Information

FHWA Toolbox for Regional Policy Analysis located at

FHWA Travel Model Improvement Program (TMIP) located at

Regression Analysis, Econometric Forecasting Techniques and Models

Econometric and statistical models are mathematical equations that can be used to describe natural and social systems. In these models, statistical techniques are used to uncover relationships or correlations between elements of these systems so that analysts can make predictions about the future. These techniques are used often in regional planning to forecast employment and population change and describe the decision-making processes of businesses, households, financial institutions, and governments.

Basic Types

There are three broad categories of statistical analysis techniques:

  • Curve Fitting/Trend Extrapolation - Trend extrapolation techniques are used to determine how one dependent variable (e.g., population, household size, or number of building permits issued) has varied with a single independent variable (time) in the past, so that a prediction may be made about the future. Spreadsheet software and statistical packages can be used to analyze time-series series data and develop best-fit curves and projections.
  • Econometric Forecasting Models – Regression and econometric techniques allow an analyst to establish the relationship between a dependent variable and one or more independent variables. For example, by establishing the correlation between past levels of employment in a particular county or city to past national economic indicators (e.g., national employment or industry output), an analyst can make predictions local activity by relying on established projections of the national indicators.
  • Discrete Choice Models – Discrete choice models can to make predictions on the probability of a decision-maker’s choice by establishing the relationship between choices and the characteristics of the decision maker (e.g., age, presence of children, number of workers, housing tenure). Information on the link between choices and the characteristics of decision-makers is often established directly through surveys (stated preference) or through observations of past behavior (revealed preference).

Typical Applications

  • No-action Future Projections - In doing an evaluation of induced growth impacts the analyst needs to compare the growth of an area's population and employment without the project (No-Action) to the future with project alternatives. Some local areas may not have estimates of future growth available at the level of detail needed (i.e., geography, time). Other areas may have a forecast that explicitly considers the effect of the proposed project in which case a projection based on current trends may be more appropriate for use as the baseline.
  • Explaining Relationships or Developing Assumptions – By establishing the relative importance of transportation among the other factors influencing past location decisions in a local area (e.g., water/sewer infrastructure, employment base, land use regulation, and ease of obtaining permits) an analyst can predict how a transportation improvement will contribute to future land use change. These types of studies can also involve quantitative evaluations of comparative cases in other regions.
  • Impacts on Property Values and Location Attractiveness – There is a growing literature in economics and planning relating changes in property values to improvements in transportation access such as interchanges and transit stops. By looking at how accessibility improvements have been capitalized into real estate prices in comparable areas, an analyst can make predictions about the effect of a proposed project on property values and ultimately land use and household or employment growth.

Strengths and Weaknesses

Econometric techniques are widely used in social science and regional planning and, when used correctly, provide an effective and defensible method on which to base conclusions about the “reasonably-foreseeable” future with our without a proposed transportation project. These techniques are often data intensive and may require considerable effort to determine if they will be useful in an evaluation. For example, an analyst may have to conduct a statistical analysis of a dataset before determining whether curve-fitting or econometric methods would produce statistically valid results. In general, econometric and statistical techniques are most applicable on large-scale systems such as regional economies, urban centers, or large corridors where large datasets can be easily obtained and individual events (e.g., business openings or closings, zoning changes) do not obscure broader trends. Although widely available desktop software packages can make the task of econometric and statistical analysis less time consuming, trained-professional judgment is required to ensure that statistical measures are accurately applied, interpreted, and summarized in documentation.

This summary was adapted from NCHRP Report 423A: Section 2, Analytical Tools.

For More Information

Council on Environmental Quality. January 1997. Considering Cumulative Effects Under the National Environmental Policy Act. Appendix A: Summaries of Cumulative Effects Analysis Methods.

Healy, J.F. 1996. Statistics: A Tool for Social Research (Fourth Edition). Wadsworth Publishing, Belmont CA.

Klosterman, R.E., R.K. Brail, and E.G. Bossard, eds. 1993. Spreadsheet Models for Urban and Regional Analysis. Center for Urban Policy Research, New Brunswick NJ.

Questions for this web site? Contact the webmaster.