TABLE OF CONTENTS

Introduction
Background
Objective
Conclusion
Scope and Methodology

Major Findings:

Finding 1 - The county’s traffic volume forecast model is widely accepted and used by state and local governments, academic institutions, and private consulting agencies.
Finding 2 - Internal control procedures could be strengthened to further assure the quality and validity of the concurrency test results.
Finding 3 - The Department of Transportation and the Puget Sound Regional Council incorporated 9 of 15 (60%) of the relevant improvements suggested by a 1996 transportation model study, and planned to incorporate the remaining improvements (40%).
Finding 4 - The Department of Transportation staff were qualified to apply and maintain the forecasting model.
Finding 5 - The county’s traffic volume forecast model was useful for roads capital improvement planning purposes.

State and countywide planning policies require transportation services and improvements to be concurrent with new development. Specifically, road improvements must be in place at the time of development, or a financial commitment made to complete improvements within six years of development approval, to maintain the level of service standards established by the county.

The Department of Transportation uses a traffic volume forecast model to measure the increased traffic volume potentially generated by proposed developments. The model forecasts are then compared to the county’s established level of service standards. If the forecasted traffic volume increases are consistent with the county’s level of service standards—roadway capacity is adequate for the estimated traffic volume increases—the Department of Transportation issues a concurrency certificate to the developer.

CONCLUSION

The conclusion is that despite the inherent weaknesses of predictive models, our conclusion was that the county’s traffic volume forecast model volume is reasonable. The model is widely accepted and used in the public and private sectors and by academic institutions, and improvements suggested by the 1996 Inside the Blackbox: Making Transportation Models Work for Livable Communities have been incorporated into the forecasting model. In addition, the Department of Transportation staff have the expertise and education to competently apply and maintain the model, and the model is useful for roads capital improvement planning purposes. However, Department of Transportation’s internal controls need to be strengthened to assure the quality and integrity of the forecasts generated by the model.

The study included a review of documents relevant to the traffic forecasting model currently used by the Department of Transportation. However, we did not examine the technical operation of the model due to lack of technical expertise. In addition, we did not test the model’s effectiveness or results in predicting the traffic impacts of development (e.g., compare predicted traffic volume to actual traffic volume), because the model reflects 1995 conditions rather than current traffic conditions. While this study reviews the model analysis based on 1995 data, it should be noted that the Department of Transportation, in cooperation with the Puget Sound Regional Council (PSRC), plans to update the forecast model later this year to reflect 1998 conditions.

Finding 1   The county’s traffic volume forecast model is widely accepted and used by state and local governments, academic institutions, and private consulting agencies.

The county’s traffic volume forecast model is actually EMME/2 software. EMME is an acronym derived from "Equilibre Multimodal, Multimodal Equilibrium," which refers to the theory of network equilibrium that underlies the multimodal travel forecast model. EMME/2 is an updated version of EMME software developed in the late 1970s at the University of Montréal.

The EMME/2 forecast model is currently used in 56 countries by over 500 organizations. In the Washington State, current EMME/2 users include the cities of Seattle, Bellevue, and Redmond; Pierce, Snohomish and Kitsap counties; the University of Washington; and several private transportation consulting firms.

Finding 2    Internal control procedures could be strengthened to further assure the quality and validity of the concurrency test results.

Internal controls procedures are established to ensure the accuracy and validity of model forecasts. Types of potential modeling errors include measurement, sampling, computational, specification, transfer, and aggregation errors. Internal controls are generally established for data input, data output, and model validation and calibration. Examples of internal control mechanisms include field observations, sensitivity analysis and expert panel review. One internal control mechanism developed by the Department of Transportation since 1996 was the creation of a review team to evaluate model results. If the review team had been established prior to the concurrency testing for the Blakely Ridge and Northridge developments, an arithmetic error made during the concurrency test would probably have been identified earlier. While the Department of Transportation has established numerous internal controls for its traffic volume forecasting process, we believe that additional enhancements, such as those suggested by the General Accounting Office, would further strengthen the quality of concurrency testing.

We recommend that the Department of Transportation 1) evaluate its internal controls for the concurrency program to incorporate the controls suggested by the General Accounting Office, the Institute of Internal Auditors, and other associations; 2) develop policies and procedures to implement the internal controls developed above; 3) and establish timelines for the implementation of the internal controls. In addition, we recommend that the department consider incorporating information regarding the results of the concurrency test and the standards used in the test on the concurrency certificate issued to the developer. Further, we recommend that the department consider incorporating sensitivity analysis to its validation and calibration process to identify variables or factors that would have a significant impact on the model output.

In 1996, a study titled Inside the Blackbox: Making Transportation Models Work for Livable Communities offered numerous suggestions to correct some common biases and problems that could affect model forecasts. Fifteen (15) of the study suggestions were relevant to the county’s forecasting model. The Department of Transportation indicated that 9 of the 15 (60%) relevant suggestions have been incorporated into the county’s model. The 6 remaining suggestions (40%) will be incorporated into the county’s model when it is updated to reflect 1998 conditions.

We recommend that Department of Transportation continue its efforts to update and improve the quality and integrity of data used in the traffic volume forecast model.

Finding 4  The Department of Transportation staff were qualified to apply and maintain the forecasting model.

Department of Transportation staff possess the experience and education necessary to apply and maintain the forecasting model competently. The 10 staff members who are responsible for the county’s forecasting model collectively possess 181 years of experience in transportation engineering; 17 PhD, master’s and bachelor’s degrees; and 2 professional engineering licenses.

Finding 5  The county’s traffic volume forecast model was useful for roads capital improvement planning purposes.