Improving ADA Complementary Paratransit Demand Estimation (2007)

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.

Improving ADA Complementary Paratransit Demand Estimation Page H-1 • Transit Cooperative Research Program Handbook for Estimating ADA Paratransit Demand Handbook for Estimating ADA Paratransit Demand Introduction The Americans with Disabilities Act of 1990 (ADA) created a requirement for comple- mentary paratransit service for all public transit agencies that provide fixed-route service. Complementary paratransit ser- vice is intended to complement the fixed- route service and serve individuals who, because of their disabilities, are unable to use the fixed-route transit system. In fulfill- ing their ADA obligations, transit operators have a responsibility to consider current and probable future demand for comple- mentary paratransit service and to plan and budget to meet all of the expected demand. The tools presented in this handbook are intended to improve transit operators’ ability to estimate the probable future demand for complementary paratransit service. In keeping with the intent of the ADA law and regulations, the methods presented are designed to predict demand for ser- vice that complies with requirements for level of service. The methods are also designed to exclude demand for ser- vices that exceed requirements for ADA complementary paratransit. Of particular importance, demand is predicted only for service by ADA-eligible individuals, for trips within three-quarters of a mile of fixed-route service, based on reservations taken from one to fourteen days in advance. Demand is predicted for service that is not capac- ity constrained by significant numbers of denials, unreliable service, or excessive telephone wait times to reach a reserva- tions agent. To the extent possible, demand is predicted only for trips that ADA-eligible individuals are unable to make by fixed- route service. The tools presented in this handbook are based on a statistical model that was esti- mated using data from 28 “representative systems” (Figure 1). The representative systems were selected from an initial list of 88 systems suggested by respondents to a survey about factors that influence the demand for paratransit. The selection process included interviews with transit agency staff, advocates, and ordinary rid- ers of each candidate system. All of the representative systems appeared to be in compliance with ADA paratransit require- ments regarding capacity constraints and generally provided quality service as of the time data were collected. Within the framework established by the ADA regulations, the representative sys- tems have a great variety of policies about on-time performance, fares, and other issues. In general, standards for service quality and users’ perceptions of service quality may vary greatly. As a result, the levels of demand estimated by the tools in this handbook are intended to correspond to realistic levels of quality service, meet- ing ADA requirements, but not necessarily meeting the expectations of all users. The demand estimation tools take account of six key variables that impact ridership. For many reasons, some variables that are known to impact demand are not included.

Page H-2 • TCRP Report 119 Improving ADA Complementary Paratransit Demand Estimation Handbook for Estimating ADA Paratransit Demand Reasons for this include lack of data, lack of reliable measures, and the small sample size that was available. The fact that a variable is not included in the demand esti- mation tools is not intended to suggest that it is not important or that transit operators should ignore it in planning for future de- mand. Despite these limitations, the tools represent a major advance in understand- ing the factors that drive demand for ADA paratransit and a major advance in transit operators’ ability to plan for the future. Overview of the Demand Estimation Tools The tools for estimating the demand for ADA complementary paratransit include: 1. An Excel spreadsheet that cal- culates demand estimates using user-entered data indicating a system’s policies and service area characteristics. 2. A series of graphs for determining factors with which demand esti- mates can be calculated by hand. 3. Elasticities and change factors for quick calculations about small differences between systems and the impacts of small changes to service policies. Figure 1 Representative Systems Haverhill- Law rence B ellingham S eattle ( K ing County) W enatchee Antioch Concord S anta Clara County Fresno Orange County S alt Lake City Denver Tulsa Fort W orth Dallas Tampa Pittsburgh S yracuse R hode I s land New York City Eugene B enton-Franklin Cincinnati B lacksburg Charlottesville Lansing Ottumw a Portland S an Mateo Co. Merrimack Vall y

Improving ADA Complementary Paratransit Demand Estimation Page H-3 • Transit Cooperative Research Program Handbook for Estimating ADA Paratransit Demand 4. A formula based on the regression model that was used to create the first three tools. 5. Tables with representative system data to use for comparison pur- poses. These tools calculate expected annual ADA paratransit ridership (including attendants and companions) when a system operates without capacity constraints as defined by the ADA regulations. The demand esti- mates are based on six factors: 1. ADA paratransit service area popu- lation. 2. Base fare for ADA paratransit. 3. Percent of applicants for ADA paratransit eligibility found condi- tionally eligible. 4. Whether or not trip-by-trip eligibility determination based on conditions of eligibility is used. 5. Percent of service area population with household incomes below the poverty line. 6. The effective window used to deter- mine on-time performance (i.e., the window from the passenger’s point of view including requirements to be ready early and adjustments made in the scheduling process that may not be communicated to passengers). How the Factors Affect Demand Briefly stated, the six factors affect demand as follows: Population: Demand increases directly in proportion to the total population of the area served. Base Fare: Demand is highly sensitive to fares, possibly even more sensitive than general public transit demand. Conditional Eligibility: Systems that have higher percentages of applicants found conditionally eligible (rather than “fully eligible” or eligible without conditions) have lower demand. Conditional Trip Determination: Systems that conduct trip-by-trip determination based on conditions of eligibility have much lower demand. Poverty Level: High levels of poverty in a service area significantly depress demand. Effective Window: Demand is highly sensitive to standards for on-time pick- ups. Systems that define “on-time” for pick-ups using a wider window have lower demand. Numerical values for these impacts, in the form of elasticities, are provided in the presentation of the third demand estimation tool. All of these factors are considered highly significant in a statistical sense. The technical report that accompanies this handbook provides additional detail about the reliability of the tools and a discussion of the mechanisms that may be responsible for the observed impacts. • • • • • •

Page H-4 • TCRP Report 119 Improving ADA Complementary Paratransit Demand Estimation Handbook for Estimating ADA Paratransit Demand Appropriate Uses of the Demand Estimation Tools As with any model, the demand estimation tools need to be used with caution. Sug- gested uses include: Planning for elimination of capacity constraints: For systems that are still experiencing difficulties with capacity con- straints, the tools provide a way of estimat- ing how much ridership may increase as these capacity constraints are removed. The calculated demand can be taken as an estimate of where growth is likely to level off, at least in the short run. In other words, the demand estimation tools provide one indication of “latent demand” in a capacity- constrained system. Benchmarking: The tools can also be used for benchmarking a system’s performance in comparison to peers. For example, if System A has ridership of 500,000 per year and System B has ridership of 750,000 in an area of twice the population, the tools pro- vide a way of comparing these two systems with adjustments for the effects of service area and service characteristics. Assessment of compliance: Comparing the demand estimate from these tools with current actual demand provides one piece of evidence about how close a paratransit system is to full compliance with the ADA requirement for no capacity constraints. Since there are many factors not included in the tools, this comparison is not conclusive. In fact, many of the representative systems used to estimate the model have ridership significantly above or below the model pre- dictions. If current demand is considerably below the level estimated by the demand estimation tools, the possibility of capacity constraints should be examined in light of other available information. Predicting the impact of policy changes: To a limited degree, the tools may also pre- dict how ridership will respond to changes in policies. However, the model’s “predic- tions” may be accurate only in the long term and might not be completely reflected in actual ridership for several years. Service planning: The impact of expand- ing or contracting the fixed-route service area (and therefore the ADA paratransit service area) can be estimated based on total population and poverty rate data for the modified service area. If predictions of population and economic conditions are available, these can be used to create long- range ridership predictions. Policy development and advocacy: By showing how sensitive paratransit demand is to various factors, the demand estimation tools can be useful in developing policies about the need for paratransit services, and for making the case for high-quality paratransit services. Cautions Policy changes within a system: Be- cause the demand estimation tools are based on a comparison of systems at one point in time, they can only be used with great caution for predicting the impact of policy changes within a system. For exam- ple, the model indicates that a system with 10% higher fares than another system will have 7% lower ridership. However, these

Improving ADA Complementary Paratransit Demand Estimation Page H-5 • Transit Cooperative Research Program Handbook for Estimating ADA Paratransit Demand differences reflect the entire history of fares at the two systems and the adjustments that riders have made to these fares over many years. In the short run, meaning one or two years, the impact of a fare change may be much less. Similar considerations apply to all of the other variables in the model. Cost management and compliance: It may be tempting to use the demand estima- tion tools as a guide to minimizing the cost of service, for example by adopting a wider pick-up window for defining on-time perfor- mance. This is not the intended use of the tools. In fact, the predictions of the model could be taken as an indication of the ex- tent to which this type of deliberate service degradation would amount to a prohibited capacity constraint, that is, a practice that limits the availability of service. These is- sues should be resolved through the public participation process at each system. Eligibility practices: The research showed a strong relationship between demand and use of trip-by-trip eligibility determination, as well as findings of conditional eligibility in the eligibility determination process. These results certainly point to the value of these tools. However, the paramount consider- ation in the eligibility process should be making use of best practices to achieve the most accurate and fairest determinations possible. Simply maximizing findings of conditional eligibility and screening out as many trips as possible would be inappropri- ate and probably illegal. Decreasing accuracy with time: The tools are based on observed demand and sys- tem characteristics in 2005 plus population data from the 2000 U.S. Census. No more recent population data were available at the time the research was conducted. It is likely that demand at the representative systems will increase over time. At a minimum, as populations grow, demand is likely to grow. In addition, it is possible that demand will grow for other reasons that are not captured in the demand estimation tools. This may be particularly true where systems have only recently eliminated denials or other capacity constraints. For these reasons, predictions from the demand estimation tools will be most meaningful within the next few years. By the time of the 2010 Census, the usefulness of the demand estimation tools will be greatly diminished. Statistical accuracy: The predictions of the demand estimation tools have a degree of inherent uncertainty. This uncertainly comes from: 1) factors that influence de- mand but were not captured in the model; and 2) the chances that the 28 “representa- tive systems” do not exactly represent the entire set of paratransit systems that are meeting ADA requirements. The statisti- cal model on which the demand estimation tools are based succeeded in explaining 96% of the observed variation in total ADA paratransit demand among the rep- resentative systems. Controlling for total population, the model explained 74% of the variation in ADA paratransit trips per capita among the representative systems. From this statistic, it is estimated that ac- tual demand should be no higher than 19% more than the predicted demand using the tools and no lower than 16% less than the predicted demand in 95% of cases. (A 95% confidence interval for the predictions is -16% to +19%.)

Page H-6 • TCRP Report 119 Improving ADA Complementary Paratransit Demand Estimation Handbook for Estimating ADA Paratransit Demand Variables Not in the Model A number of factors commonly believed to influence demand for paratransit are not in the demand estimation tools. Some of the notable cases include: Population in older age groups: The research found that the percentage of the population that is above the age of 65 or 75 did not significantly affect paratransit de- mand at the representative systems. This outcome may reflect the fact that younger people with disabilities ride more frequently than older people. As a result, even though older people tend to account for a majority of ADA eligible people, they do not neces- sarily account for a majority of demand. The model result could also, at least in part, stem from limitations of ADA paratransit from the perspective of older people. Incidence of disability: Census data indicate that the percentage of the popula- tion with a disability varies greatly among metropolitan areas. However, the research found no statistically significant relationship between paratransit demand and Census measures of the population with a dis- ability. This may be a result of the fact that none of the questions about disability in the Census measures ability to use public transportation. Availability of human service transpor- tation: The availability of human service transportation almost certainly has a major impact on ADA paratransit demand. An attempt was made to measure the overall availability of human service transportation at the representative systems. However, this effort produced only partial and inexact results that were not statistically related to ADA paratransit demand. The absence of a factor related to human service transporta- tion is a limitation of the demand estimation tools that users should address through knowledge of local conditions. Availability and quality of accessible fixed-route transit: It is widely assumed that high levels of accessible transit service or high levels of transit service in general will reduce the demand for ADA paratransit. However, the research did not find a sta- tistically significant relationship between paratransit demand and availability of accessible transit or availability of transit service overall. In fact, contrary to expecta- tions, the research showed that paratransit demand may be higher in places that have extensive transit service (including acces- sible transit service) than in places with less extensive transit service. This topic is ad- dressed at length in the technical report. Telephone access: Difficulty getting through on the telephone to make a reser- vation almost certainly affects paratransit demand. An attempt was made to capture this effect by requesting data about tele- phone hold times. However, nine of the 28 representative systems were not able to provide a quantitative measure of telephone hold time. As a result of this data limita- tion, the observed relationship between hold times and demand was not statisti- cally significant, although it was nearly so and in the expected direction. The techni- cal report provides more detail. Systems where customers face long hold times or

Improving ADA Complementary Paratransit Demand Estimation Page H-7 • Transit Cooperative Research Program Handbook for Estimating ADA Paratransit Demand frequent busy signals should assume that remedying this situation may well result in higher demand levels (other factors being equal) even though the demand estimation tools do not provide a quantitative estimate of this effect. Ethnicity and language: It is possible that certain ethnic groups may use paratransit less than others because of traditions about taking care of family members. The research did not find any statistically sig- nificant impact, but did not rule it out. In communities where numerous languages are spoken, lack of marketing and mul- tilingual reservations staff may reduce paratransit demand. Language issues were not investigated. As communities become more diverse, these issues may become particularly important in some areas. Policy and Planning Implications The demand estimation tools may be useful for developing policies and plans for the future. Issues that may be informed by the research include how demand will grow in the future and how systems’ policies limit demand. Long-term demand growth: The re- search results imply that demand will grow in proportion to total population and is not related to the proportion of the population in older age groups. If this result is correct, the anticipated graying of America may have much less impact on ADA paratransit demand than expected. The result could also indicate that responding to needs of older people will require developing solu- tions other than ADA paratransit. Eligibility: The research finding that conditional eligibility and use of trip-by-trip eligibility determination have significant impacts on demand points to a need for continued work to provide transit opera- tors with the best possible eligibility as- sessment tools. The widespread adoption of functional assessment for eligibil- ity determination has aroused concern among some in the disability community about the fairness and accuracy of the implementation of these methods in some systems. The state of the art with respect to trip-by-trip determination is still very rudimentary. The results of this research suggest that trip-by-trip determination has a much greater impact on demand than previously suspected. This points to an urgent need to spread the use of existing best practices and to improve the state of the art in this area. On-time performance: While ADA regu- lations prohibit “substantial numbers of significantly untimely pick-ups” as one type of capacity constraint that “limits the availability of complementary paratransit service” (49 CFR 37.131(f)), standards for what amounts to an untimely pick-up vary among systems. The findings of the research about how these standards im- pact demand may be useful in formulating policy about the point at which overly loose on-time standards begin to limit the avail- ability of service. Economic conditions: The research showed that high levels of poverty in a com- munity depress demand. Communities that are able to raise overall standards of living will mostly like see an increase in demand

Page H-8 • TCRP Report 119 Improving ADA Complementary Paratransit Demand Estimation Handbook for Estimating ADA Paratransit Demand for paratransit services. Unfortunately, the research was not able to identify the likely impact of improving the economic condition of people with disabilities. Fixed-route transit and paratransit demand: The research did not find any tendency for high levels of fixed-route transit service (including accessible transit service) to reduce paratransit demand. This tentative result suggests a need for further research about how people with disabilities make choices regarding how they travel. Instructions for Using the Demand Estimation Tools Spreadsheet Tool An Excel spreadsheet is provided that cal- culates expected annual ADA paratransit ridership per capita and total ridership when a system operates without capacity con- straints as defined by the ADA regulations. The spreadsheet can be downloaded from the TCRP website along with the electronic version of this report. Pop-up instructions provide guidance about how to enter vari- ables where there could be confusion. Cells that require values in specific ranges (i.e., 0 to 100 for percentages, and 0 or 1 for con- ditional trip determination) have validation rules that prevent other values from being entered. Figure 2 shows the tool with pop- up instructions for a sample cell. The inputs needed to use the spreadsheet are as follows: ADA service area population = total population according to the 2000 U.S. • Census for the actual area served by ADA paratransit. Depending on service policies, this may be just the area three- quarters of a mile around fixed-route service or a larger area. It is critical that the actual ADA service area be used, or an area as close as possible to the actual ADA service area. Base Fare = the full cash fare for an ADA paratransit trip before any discounts for advance purchase or use of a monthly pass, and before adding any zone charges. The percent of applicants found conditionally eligible = 100 x (the number of people found eligible with conditions) ÷ (the number of people who apply for ADA paratransit eligibility). The most recent full year of eligibility statistics should be used. Conditional trip determination = 1 if trip-by-trip determination based on conditions of eligibility is done, 0 otherwise. Percent below the poverty rate = 100 x (the number of people in households with incomes below the poverty rate in the area actually served by ADA paratransit as reported in the 2000 U.S. census) ÷ (the ADA service area population from the first bullet). Effective on-time window = the total variation in pick-up time, before or after the last time that was given to the customer, before the trip is no longer counted as being “on-time.” For example, if a vehicle is considered late beginning 20 minutes after the promised time, but customers are expected to be ready 10 minutes before the promised time, then the “effective window” is 30 minutes. Similarly, if pick-up times can be changed by up to 10 minutes without informing the customer, then the effective window may need to be adjusted. • • • • •

Improving ADA Complementary Paratransit Demand Estimation Page H-9 • Transit Cooperative Research Program Handbook for Estimating ADA Paratransit Demand The spreadsheet gives predicted annual ridership and annual ridership per capita, as well as confidence limits for these. A sepa- rate tab provides a graphical representation of the confidence limits. The spreadsheet also includes data from the representative systems used to develop the demand estimation tool. Information for each system includes service characteris- tics, measures of service quality, eligibility statistics, and area demographics, Users can use this data to look for systems that are similar to their own, or to explore the possible influence of variables that were not included in the demand estimation tool itself. Figure 2 Spreadsheet Tool for Estimating ADA Paratransit Demand

Page H-10 • TCRP Report 119 Improving ADA Complementary Paratransit Demand Estimation Handbook for Estimating ADA Paratransit Demand Graphical Demand Estimation Tool A form is given in Figure 3 that can be used for hand calculations. There is one row for each factor of the estimation tool. In the boxes to the right, enter values for each fac- tor by reading from the graphs in Figures 5 through 8. For each graph, users locate the value along the horizontal axis that applies to their system and read the factor from the vertical axis. The inputs needed to use the graphs are the same ones described for the spreadsheet tool. A worked-out example is provided in Figure 4, using the same input values illustrated for the spreadsheet tool in Figure 2. In the first row, the service area population is entered, rounded to three significant figures. In the second row, the constant of 31.91 is carried over. In the third row, the fare of $2.00 is located on the horizontal scale of the “Factor for Base Fare” graph, then a line is traced up to the curve and read across to the vertical scale, giving a factor of 0.59. The process is repeated for the remaining factors. The result is rounded to three significant f igures as 139,000 annual trips, approximately matching the result from the spreadsheet tool. • • • • • Figure 3 Calculation Form for Use with Graphical Tools Total ADA Service Area Population X X 31.91 X X Base Fare Factor X X Eligibility Factor X X Conditional Trip Screening Factor: 0.52 if trips are screened 1.0 if trips are not screened X X Poverty Factor X X On-time Window Factor = = ADA Paratransit Trips per Year (Including Attendants and Companions)

Improving ADA Complementary Paratransit Demand Estimation Page H-11 • Transit Cooperative Research Program Handbook for Estimating ADA Paratransit Demand Figure 4 Example Calculation with Graphical Tools Total ADA Service Area Population X X 31.91 X X Fare Factor X X Eligibility Factor X X Conditional Trip Screening Factor: 0.52 if trips are screened 1.0 if trips are not screened X X Poverty Factor X X On-time Window Factor = = 448,000 31.91 0.59 0.84 0.52 0.39 0.098 ADA Paratransit Trips per Year (Including Attendants and Companions) 139,000 0 0.5 1 1.5 $0.00 $0.50 $1.00 $1.50 $2.00 $2.50 $3.00 Base Fare B as e Fa re F ac to r

Page H-12 • TCRP Report 119 Improving ADA Complementary Paratransit Demand Estimation Handbook for Estimating ADA Paratransit Demand Figure 5 Factor for Base Paratransit Fare 0 0.5 1 1.5 2 2.5 3 $0.00 $0.50 $1.00 $1.50 $2.00 $2.50 $3.00 $3.50 $4.00 $4.50 $5.00 Base Fare B as e Fa re F ac to r

Improving ADA Complementary Paratransit Demand Estimation Page H-13 • Transit Cooperative Research Program Handbook for Estimating ADA Paratransit Demand Figure 6 Factor for Percent Found Conditionally Eligible 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Percent Conditionally Eligible El ig ib ili ty F ac to r

Page H-14 • TCRP Report 119 Improving ADA Complementary Paratransit Demand Estimation Handbook for Estimating ADA Paratransit Demand Figure 7 Factor for Percent Below Poverty Level 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0% 5% 10% 15% 20% 25% 30% 35% 40% 45% 50% Percent of Population below Poverty Level Po ve rt y Fa ct or

Improving ADA Complementary Paratransit Demand Estimation Page H-15 • Transit Cooperative Research Program Handbook for Estimating ADA Paratransit Demand Figure 8 Factor for On-time Window 0.05 0.06 0.07 0.08 0.09 0.1 0.11 0.12 0.13 0.14 0.15 0.16 0.17 0.18 0.19 0.2 10 15 20 25 30 35 40 45 50 55 60 Effective Window (Minutes) O n- tim e W in do w F ac to r

Page H-16 • TCRP Report 119 Improving ADA Complementary Paratransit Demand Estimation Handbook for Estimating ADA Paratransit Demand Elasticities and Difference Factors The demand model provides elasticities for some variables and “difference factors” that function in a similar way for others. These can be used to help compare two systems or, in some cases, to estimate the effect of small changes. Figure 9 shows how these factors apply to differences between systems or changes of 1%. However, ap- plying these factors to differences much greater than 1% requires application of exponentials. It is recommended that users consult the graphs in Figures 5 through 8. For example, ridership with a $1.50 fare and Figure 9 Elasticities and Different Factors ridership with a $2.00 fare can be compared as follows: From Figure 4, the factor for a base fare of $1.50 is 0.73. From the same figure, the factor for a base fare of $2.00 is 0.59. All else being equal, a system with a $2.00 base fare would be expected to have rider- ship 0.59/0.73 = 0.81 times that of a system with a $1.50 base fare. Variable Elasticity Factor Interpretation Base Factor -0.77 A 1% higher base fare (e.g., $2.02 vs. $2.00) corresponds to 0.77% less demand. Percent Conditionally Eligible -0.29 at the mean A 1% higher percent found conditionally eligible compared to the mean value of 21% (21.21% vs. 21%) corresponds to 0.29% less demand. 1.39 A 1% greater percentage of applicants found conditionally eligible (e.g., 31% vs. 30%) corresponds to 1.39% less demand. Conditional Trip Screening 48% Systems that use conditional trip screen- ing have 48% lower demand than other systems. Percent below Poverty -0.90 at the mean A 1% higher poverty rate compared to the mean value of 13% (13.13% vs. 13%) corresponds to 0.90% less demand. -6.6 A 1% higher percentage of the population below the poverty level (e.g., 16% vs. 15%) corresponds to 6.6% less demand. Effective Window -0.72 A 1% wider effective window (e.g., 30.3 min- utes vs. 30 minutes) corresponds to 0.72% less demand.

Improving ADA Complementary Paratransit Demand Estimation Page H-17 • Transit Cooperative Research Program Handbook for Estimating ADA Paratransit Demand Figure 10 Formula for Predicting Demand Formula-Based Estimation For those who are comfortable with math- ematics, a formula is provided that is the basis of the other tools. Most users will probably prefer to use the graphical tools, the spreadsheet provided with the hand- book, or the elasticities and difference factors. Based on the experience of 28 represen- tative systems, a formula that predicts de- mand for ADA complementary paratransit trips as of 2005 is given in Figure 10. In this formula, “exp” refers to exponentiation, that is, “e” (the base of the natural logarithms) raised to the power of the term in parenthesis. All of the population data should be from the 2000 U.S. Census. All of the population data should be for the actual area served by ADA paratransit. Depending on service policies, this may be just the area three- quarters of a mile around fixed-route service or a larger area. It is critical that the actual ADA service area be used, or an area as close as possible to the actual ADA service area. Base Fare = the full cash fare for an ADA paratransit trip before any discounts for advance purchase or use • • • • of a monthly pass, and before adding any zone charges. The percent of applicants found conditionally eligible = 100 x (the number of people found eligible with conditions) ÷ (the number of people who apply for ADA paratransit eligibility). The most recent full year of eligibility statistics should be used. Conditional trip determination = 1 if trip-by-trip determination based on conditions of eligibility is done, 0 otherwise. Percentage below the poverty rate = 100 x (the number of people in households with incomes below the poverty rate in the area actually served by ADA paratransit as reported in the 2000 U.S. census) ÷ (the ADA service area population). Effective On-time Window = the total variation in pick-up time, before or after the last time that was given to the customer, before the trip is no longer counted as being “on-time.” For example, if a vehicle is considered late beginning 20 minutes after the promised time, but customers are expected to be ready 10 minutes before the promised time, then the “effective window” is 30 minutes. Similarly, if pick-up times can be changed by up to 10 minutes without informing the customer, then the effective window may need to be adjusted. • • • • ADA Paratransit Trips per Year = (Total ADA Service Area Population) x 3.463 x (Base Fare)-0.772 x exp (1.385 x (Percent of Applicants Found Conditionally Eligible/100)) x exp (-0.662 x (Conditional Trip Determination)) x exp (-6.633 x (Percent of Population below Poverty/100)) x (Effective On-time Window)-0.722