Passenger Demand Forecasting

(1)

February 2014

Passenger Demand

Forecasting

(FY2014 ─ FY2018)

HDR Engineering, Inc. 8403 Colesville Road Suite 910 Silver Spring, MD 20910

Final Report

(2)

PASSENGER DEMAND FORECASTING

LIST OF TABLES

Table 1: Passenger Trip Requests by Region (FY2009 – FY2013) ... 11

Table 2: Ridership by Service Region (FY2009 – FY2013) ... 14

Table 3: Sample of Performance Metrics Published in Annual/ Monthly Reports ... 25

Table 4: Variables Tested in Regression Analysis ... 39

Table 5: Regression Results ‒ Eastern ... 41

Table 6: Regression Results ‒ Antelope Valley ... 42

Table 7: Regression Results ‒ Northern ... 43

Table 8: Regression Results ‒ Southern ... 44

Table 9: Regression Results ‒ Santa Clarita ... 45

Table 10: Regression Results ‒ West/ Central ... 46

Table 11: Current Fare Structure ... 48

Table 12: Board Approved Nominal Base Fares (FY2014 – FY2018) ... 49

Table 13: Inflation Projections (FY2014 – FY2018) ... 49

Table 14: Free Fare Projections (FY2014 – FY2018) ... 49

Table 15: New Applicants (FY2014 – FY2018)... 50

Table 16: Santa Clarita Eligibility Evaluations (FY2014 – FY2018) ... 50

Table 17: West/ Central Eligibility Evaluations (FY2014 – FY2018) ... 50

Table 18: Real Gasoline Price Change Projections (FY2014 – FY2018) ... 51

Table 19: Unemployment Level Projections (FY2014 – FY2018) ... 51

Table 20: Operations Forecasts, Mean Expected Outcome (FY2014 – FY2018) ... 52

Table 21: Ridership Forecasts by Service Region, Mean Expected Outcome (FY2014 – FY2018) ... 52

Table 22: Steady State Scenario (FY2014 – FY2018) ... 54

Table 23: Risk-Adjusted Monthly Passenger Trips Completed Forecast (FY2016) ... 57

Table 24: Current Fare Structure ... 58

Table 25: Access Services Operations Forecasts, Mean Expected Outcome (FY2014 – FY2018) ... 59

Table 26: Risk-Adjusted Ridership Projections (FY2016 – FY2018) ... 59

Table 27: Regression Results – Service Area ... 63

Table 28: Regression Results ‒ Eastern ... 64

Table 29: Regression Results ‒ Antelope Valley ... 65

Table 30: Regression Results ‒ Northern ... 66

Table 31: Regression Results ‒ Southern ... 67

Table 32: Regression Results ‒ Santa Clarita ... 68

Table 33: Regression Results ‒ West/ Central ... 69

Table 34: New Applicant Forecasts by Service Region, Mean Expected Outcomes (FY2014 – FY2018) 70 Table 35: Risk-Adjusted New Applicant Projections (FY2016 – FY2018) ... 70

LIST OF FIGURES

Figure 1: Monthly Trip Requests (FY2004 – FY2013) ... 11

Figure 2: Seasonality of Paratransit Demand (FY2009 – FY2013) ... 12

Figure 3: Ridership in Service Area (FY2004 – FY2013) ... 13

Figure 4: Ridership in the Four LA Basin Regions (FY2004 – FY2013) ... 15

Figure 5: Distribution of Passengers by Service Region (FY2009 – FY2013) ... 16

Figure 6: Ridership by Type of Passenger (FY2004 – FY2013) ... 17

Figure 7: Trip Requests and Average Real Fare (FY2004 – FY2013) ... 18

Figure 8: Ridership and Complaint Rate (FY2004 – FY2013) ... 19

Figure 9: Ridership and Complaint Rate by Service Region (FY2009 – FY2013)... 20

(5)

PASSENGER DEMAND FORECASTING

Figure 11: Trip Requests and No-Show and Cancellation Rates (FY2004 – FY2013) ... 22

Figure 12: Population Growth in Los Angeles County (2003 – 2018) ... 23

Figure 13: Senior Population Growth in Los Angeles County (2003 – 2018) ... 23

Figure 14: National Peer Review, Number of Passenger Trips (FY2008 – FY2011) ... 28

Figure 15: Regional Peer Review, Number of Passenger Trips (FY2008 – FY2011) ... 29

Figure 16: National Peer Systems, No Show Rates (FY2008 – FY2011) ... 29

Figure 17: National Peer Systems, Late Cancellation Rates (FY2008 – FY2011) ... 30

Figure 18: National Peer Review, Real Operating Cost per Passenger Trip (FY2008 – FY2011) ... 31

Figure 19: Regional Peer Review, Operating Cost per Passenger Trip (FY2008 – FY2011) ... 31

Figure 20: National Peer Review, Passengers per Revenue Hour (FY2008 – FY2011) ... 32

Figure 21: Regional Peer Review, Passengers per Revenue Hour (FY2008 – FY2011) ... 33

Figure 22: National Peer Review, Farebox Recovery (FY2008 – FY2011) ... 34

Figure 23: Regional Peer Review, Farebox Recovery (FY2008 – FY2011) ... 34

Figure 24: National Peer Review, On Time Performance (FY2008 – FY2011) ... 35

Figure 25: Regional Peer Review, On Time Performance (FY2008 – FY2011) ... 35

Figure 26: Structure and Logic Diagram of the Paratransit Demand Model ... 37

Figure 27: Trip Requests by Service Region, Actual Data vs. Backcast (Jul. 2012 – Jun. 2013) ... 47

Figure 28: Ridership by Service Region (FY2008 – FY2018) ... 53

Figure 29: Trip Requests and Steady State Scenario (FY2014 – FY2018) ... 55

Figure 30: Risk Analysis of Paratransit Demand Forecast ... 56

Figure 31: Forecast of Passengers (FY2014 – FY2018) ... 57

Figure 32: Probability Distribution of Passengers, in Millions (FY2014- FY2016) ... 58

Figure 33: Trip Requests and Eligibility Evaluations (FY2008 – FY2013) ... 60

(6)

PASSENGER DEMAND FORECASTING

Executive Summary

HDR Decision Economics (HDR) has been retained by Access Services (Access) to provide an updated paratransit demand analysis for fiscal year (FY) 2014 to 2018. Since 2004, HDR has been providing historical demand analysis and forecasting services to Access; in this update, a new methodology for developing paratransit forecast for each service region is employed. Part of the update also includes two new analyses: analysis of peer systems and analysis of new applicants.

Peer Analysis

The goal of a peer review is to better understand an agency’s strengths and weaknesses so as to formulate strategies to improve its performance. For Access, the review involves examining national and local paratransit agencies over four fiscal years in five areas of performance: service utilization (ridership), cost efficiency (operating cost per trip), productivity (number of passengers per revenue hour), cost effectiveness (farebox recovery), and service quality (one-time performance). The analysis can serve as a valuable management tool designed to help improve Access’s service and operation performance. In terms of service utilization, Access averages about 2.8 million passenger trips a year, an estimate that is well above the national peer systems median value of 1.6 million. The average annual ridership growth for Access is 5.5 percent, which is the fourth strongest, behind agencies in Chicago, Washington D.C. and Boston. When comparing cost efficiency, a disparity in average annual operating costs between national paratransit systems is found. Paratransit agencies in Chicago, Washington and Boston all have average operating costs over 40 dollars per trip while agencies in Philadelphia, Houston and Miami have average costs under 30 dollars per trip. The finding may be due to differences in employee compensation across different systems. For Access, operating cost per trip is about equal to the national peer systems median value of $34.3.

Access is one of the more productive paratransit agencies in terms of number of passengers per revenue hour. It averages 1.8 passengers per revenue hour which is higher than all of the national peer systems except SEPTA in Philadelphia (SEPTA also averages around 1.8 passengers). The median value for all national peer systems is 1.7 passengers per hour. In terms of cost effectiveness, Access has an average farebox recovery ratio of 5.2 percent, which is on par with the median value of other national peer systems. Access is however less effective than three other peer systems, which report farebox recovery ratios between 6 and 13 percent. The last performance measure compares service quality using on-time performance, but this may be biased as agencies have different definitions of what is considered on-time. The reported data show that the percentage of trips completed on time for Access Service is on par with the median value for national agencies, averaging about 91 percent from 2008 to 2011. However the data are far below Boston, which averages over 98 percent annually.

Overall the analysis finds that Access compares well with peer agencies across the country. In all five areas of performance analyzed, Access continues to show improvement year after year. The results indicate that Access may be more resilient in anticipating and handling potential operational and demand changes than other peers.

Demand Analysis

Trip demand is estimated through completed trips, which are derived from statistical and econometric analyses of trip requests. The analysis framework involves statistical methods for studying historical trends and econometric techniques for determining factors that drive paratransit demand. The combined analysis leads to a series of econometric regressions that quantify the causality of market dynamics in paratransit demand. These dynamics are examined for each service region using monthly operating data and other socio-economic data (unemployment, population, etc.) from federal, state, and local agencies. Date-specific dummy variables are used to capture specific events that influence paratransit demand. There are six regression models, one for each region. Service region-specific models are independent of each other as interdependency due to operational events is captured by date-specific dummy variables.

(7)

PASSENGER DEMAND FORECASTING

Each model is also independent of qualitative service quality (such as customer satisfaction), alternative transportation modes available, and general travel demand patterns within the regions.

The historical data show that the number of trips completed is closely related to the number of trip

requests and both have experienced similar cyclical patterns over the past decade. In particular, ridership increased in all regions in FY2008 and FY2009, before decreasing in FY2010 as the economy recovered from the recession. Since 2010, ridership has steadily increased and reached its highest level in FY2013. Over the last three years, overall ridership has grown by an average annual rate of 7.9 percent.

Ridership forecasts are derived from passenger trip requests based on the average completion rate observed at the service area level in FY2012. Cancellations and no-shows are derived in the same way. Note that a 0 percent denial rate is assumed throughout the forecasting period. Operations forecasts are summarized in Table ES 1.

Table ES 1: Operations Forecasts, Mean Expected Outcome (FY2014 – FY2018)

Results 2014 2015 2016 2017 2018

Passenger Trip Requests 3,946,288 4,156,981 4,379,282 4,627,156 4,876,274

% Change 9.9% 5.3% 5.3% 5.7% 5.4%

Cancellations 70,204 74,611 78,600 83,049 87,521

No-Shows 80,774 85,845 90,435 95,554 100,698

Passengers 3,795,310 3,996,526 4,210,246 4,448,553 4,688,055

% Change 9.0% 5.3% 5.3% 5.7% 5.4%

Note: FY2014 projections include actual estimates through December 2013.

The number of trip requests is expected to grow above one million over the forecasting period and is expected to come close to the five million mark in FY2018. The increase translates to a 35.8 percent growth in five years (from FY2013). Passenger trips completed are expected to rise by nine percent in FY2014. Growth rates for the subsequent four fiscal years are significantly lower, mainly because of growth deceleration in Southern, Eastern, and West/ Central regions. Projections by service region are presented in Table ES 2.

Table ES 2: Ridership Forecasts by Service Region, Mean Expected Outcomes

(FY2014 – FY2018)

Fiscal

Year TOTAL Eastern

West/

Central Northern Southern

Antelope Valley Santa Clarita Backup 2013 3,481,204 977,840 540,810 668,668 1,131,881 111,263 46,381 4,361 2014 3,795,310 1,040,337 577,717 717,239 1,265,699 146,314 43,469 4,535 9.0% 6.4% 6.8% 7.3% 11.8% 31.5% -6.3% 4.0% 2015 3,996,526 1,099,936 594,250 755,236 1,312,206 184,180 45,872 4,847 5.3% 5.7% 2.9% 5.3% 3.7% 25.9% 5.5% 6.9% 2016 4,210,246 1,144,561 620,928 820,741 1,350,670 220,567 47,673 5,106 5.3% 4.1% 4.5% 8.7% 2.9% 19.8% 3.9% 5.3% 2017 4,448,553 1,196,098 649,940 894,264 1,397,474 256,002 49,378 5,395 5.7% 4.5% 4.7% 9.0% 3.5% 16.1% 3.6% 5.7% 2018 4,688,055 1,232,652 681,072 974,173 1,447,422 295,906 51,146 5,685 5.4% 3.1% 4.8% 8.9% 3.6% 15.6% 3.6% 5.4%

(8)

PASSENGER DEMAND FORECASTING

Analysis of New Applicants

HDR has also investigated the rapid growth in new applicants since 2009 by means of a statistical analysis. HDR’s approach aims at integrating the analysis of new applicants into the demand analysis framework. The regional trends in trip requests and eligibility evaluations show that since FY2010, evaluations have been growing almost as much as trip requests. Data on service region eligibility evaluations shed light on potential reasons for spikes in FY2010. While eligibility evaluations and trip requests are highly correlated, they may not be so for certain regions such as West/ Central and Santa Clarita. Moreover, the volatility in eligibility evaluations, and ultimately in the new applicant data, may reflect the need for dummy variables to capture service region-specific events.

The analysis shows that service region-specific new applicants are very likely to be explained by qualitative service factors that are not captured by the data at hand. To track deviations in historical trends due to operational changes, date-specific dummy variables are used in multiple occasions. Given that service region-specific models are limited to a pure time series framework, an additional aggregate model is developed to investigate how new applicants at the service area level can be impacted by variables used in the demand analysis.

The modeling results suggest that the number of new applicants can be explained by real gas price and unemployment (lagged one month), which reflects the impact of the 2008-09 economic recession and the residual uncertainty on the U.S. economy. Since the recession, many transit agencies have been facing funding challenges and paratransit services have been negatively impacted. As a result, riders, especially those from areas with low-income and/ or rapidly growing Asian and Latino populations such as the Southern region, may have diverted trips to Access, thus creating a surge in new applications and eligibility evaluations. The results also suggest that new applicant data are sensitive to operational changes. Dummy variables for specific event dates show that changes in service region boundaries and installation of reservation/ scheduling software may be influential.

New applicant projections are provided in Table ES 3. The annual forecasts suggest that aside from Southern region and Antelope Valley, new applicant trends are likely to align with those of ridership in each service region. The results suggest that socioeconomic factors such as rapidly aging population, income, and unemployment may continue to be significantly impacting the forecasts in those regions.

Table ES 3: New Applicant Forecasts by Service Region, Mean Expected

Outcomes (FY2014 – FY2018)

Fiscal

Year TOTAL Eastern

West/

Central Northern Southern

Antelope Valley Santa Clarita 2014 29,676 6,874 5,425 3,335 11,760 2,114 167 10.7% 7.2% 10.5% 6.7% 10.6% 38.0% -16.4% 2015 30,168 6,226 5,342 3,213 12,574 2,672 142 1.7% -9.4% -1.5% -3.7% 6.9% 26.4% -15.4% 2016 33,824 6,273 5,732 3,519 14,678 3,475 148 12.1% 0.8% 7.3% 9.5% 16.7% 30.1% 4.4% 2017 38,113 6,320 6,150 3,855 17,112 4,522 154 12.7% 0.8% 7.3% 9.5% 16.6% 30.1% 4.5% 2018 43,187 6,367 6,599 4,222 19,953 5,884 161 13.3% 0.8% 7.3% 9.5% 16.6% 30.1% 4.5%

Note: FY2014 projections include new applicant estimates derived from actual observations of eligibility evaluations through January 2014.

(9)

1.

Introduction

Access Services (“Access”), a local governmental agency created in 1994, is the Los Angeles County Consolidated Transportation Services Agency (CTSA) that provides Americans with Disabilities Act (ADA) mandated paratransit service for eligible persons in Los Angeles County. Access is available to any location within ¾ of a mile of any public bus fixed route and within ¾ of a mile around METRO Rail stations during operating hours. The service area covered by Access is divided into six regions and extends into portions of the surrounding counties of San Bernardino, Orange and Ventura.

HDR Decision Economics (HDR) has been providing paratransit demand analysis and forecast to Access since 2003 and was recently retained to provide an update.

Objectives of the Study

The paratransit demand analysis relies on historical data and forms the basis for future projections. It involves detailed and scientific examination, both at the system and regional levels, of trends and movements in trip demand and its constitutive elements such as cancellations, no-shows, missed trips, and trips completed.

More specifically, the key analytical tasks involve:

 Examining the behavior of trip demand over time in relation to both internal operational changes to Access policies (e.g., new fare structure) and external modeling and socio-economic factors (e.g., seasonality and fluctuations in fuel prices);

 Identifying potential structural breaks in the data series (caused by changes in market conditions for instance); and

 Estimating the degree of correlation among different variables (such as trip requests and population).

HDR is building upon its database of Access operational statistics, which has been continuously

maintained since 2003. The database includes monthly operating and financial data at the regional level for the last eighteen years. As part of the analysis update, HDR has reviewed the updated database and validated the sampling methods used by Access to produce some of the trip demand and performance measures used in the analysis.

Similar to the annual studies conducted in the past, HDR has assembled historical demographic and socio-economic data (population by age group, employment, retail gasoline prices, local consumer price index, etc.) from various state and national sources such as the California Department of Finance, the United States (U.S.) Census Bureau, the Bureau of Labor Statistics (BLS), and the Energy Information Administration (EIA).

There are two new tasks for this update. The first is a peer review analysis. The peer review is a high-level analysis that draws, in part, from previous HDR projects from large and small agencies such as Washington Metropolitan Area Transit Authority (WMATA) and Riverside Transit Agency (RTA). Additional data come from peer agencies, Federal Transit Administration’s National Transit Database (NTD), Florida Transit Information System (FTIS), New York City Transit’s Paratransit Peer Reports, and agency operation and service annual reports. The ultimate objective of the peer review is to identify demand-related issues (increase in customer complaints, high no-show rate and transfer of ridership from other specialized service providers, etc.) that have arisen elsewhere and examine how these issues have been addressed.

The second new task is a time-series econometric analysis and forecast development of new applicants. The purpose of the analysis is to investigate the possible causes of the rapid increase in new applicants since 2009. The results of the analysis are expected to help Access better anticipate the impacts of variations in new applicants on its paratransit operations.

(10)

PASSENGER DEMAND FORECASTING

Plan of the Report

The report includes full technical documentation of the model, including historical data, analytical framework, specification experiments and diagnostic tests, forecasting assumptions and any policy scenarios investigated. Following this introduction, a historical overview of key operating measures of Access paratransit trip demand is presented in Section 2. The summary of operations leads to a discussion of performance metrics in Section 3 and the performance-based peer analysis in Section 4. Section 5 describes the demand analysis framework and resulting demand outcomes while Section 6 reports forecasting assumptions and results, as well as those under an alternate fare scenario. The report concludes with the analysis of new applicants in Section 7.

The report also contains a number of appendices. A list of all acronyms used in the report is provided in Appendix 1. A glossary of all technical terms used in the report is provided in Appendix 2 to further explain the methodology and interpretation of the results. A risk analysis primer is included in Appendix 3. Monthly ridership and new applicant projections are provided for each region served by Access in

appendices 4 through 6. Appendix 7 contains a map of the service area. All data sources and references used throughout the study are listed in Appendix 8.

(11)

PASSENGER DEMAND FORECASTING

2.

Historical Overview

This section presents a historical overview of paratransit operations data for the six regions served by Access Services from July 2008 to June 2013. The six regions include Eastern, Northern, Southern, West/ Central, Santa Clarita and Antelope Valley. Unless otherwise noted, the discussion pertains to fiscal year (FY) rather than calendar year. The overview is supported by the analysis of the main factors shaping trip demand for Access.

Trip Demand

Passenger trips requested and ridership are used as indicators of the demand for paratransit service. Passenger trip requests include all trips completed, no-shows, cancellations and trips denied. Ridership refers to passenger trips completed.

Trip Requests

Passenger trip requests in Access’s entire service area grew to 3.61 million in FY2013, from 2.98 million in FY2009 – at an annual rate of 5.3 percent. From 2004 to 2007, trip requests declined. As the U.S. economy recovered after the 2008-09 recession, Access experienced steady growth in the number of trip requests, with an average annual growth of 5.7 percent since 2008. During that period, 2010 was the only year with negative growth, which can be explained by an increase in fares and the dropping of a

subcontractor in the Southern and West/ Central regions. Since 2010, the number of annual trips requested has increased by no less than 23 percent.

Overall, trip demand increased in each of the previous three years in every region of Los Angeles County. In 2013, the largest regions in Los Angeles County in terms of trip requests were the Eastern and

Southern regions. Since 2010, the Eastern and Southern regions have accounted for 57 percent of all growth in trip requests. Across all regions, West/ Central is the only one with fewer trip requests in FY2013 than in FY2002. This reflects the impact of the change in regional boundaries in September 2006 and September 2007, when portions of the West/ Central region were transferred to the Southern region. Additionally, a change in contractor in November 2009 led to a 6.5 percent drop in the trip requests in FY 2010; this marked the largest decline among all regions in two years.

These demand and growth estimates are also reported in Table 1, along with trip requests for “backup”, an around-the-clock service provided by Access Services in case of failure of the carrier (e.g., the vehicle has not arrived by the scheduled pick up time plus the 20-minute on-time window). Figure 1 shows monthly trip requests for the whole service area from July 2003 to July 2013.

(12)

PASSENGER DEMAND FORECASTING

Table 1: Passenger Trip Requests by Region (FY2009 – FY2013)

FY2009 FY2010 FY2011 FY2012 FY2013

TOTAL 2,983,144 2,928,724 3,122,945 3,419,049 3,609,044 10.0% -1.8% 6.6% 9.5% 5.6% Antelope Valley 43,631 69,561 76,413 86,906 116,025 13.0% 59.4% 9.8% 13.7% 33.5% Eastern 886,589 881,630 919,267 975,921 1,005,145 5.2% -0.6% 4.3% 6.2% 3.0% Northern 542,235 546,526 597,855 665,517 691,179 10.6% 0.8% 9.4% 11.3% 3.9% Santa Clarita 36,319 48,584 56,326 58,523 60,956 32.9% 33.8% 15.9% 3.9% 4.2% Southern 966,673 907,250 942,694 1,077,575 1,171,582 18.9% -6.1% 3.9% 14.3% 8.7% West/ Central 502,555 470,038 525,472 549,355 559,336 1.6% -6.5% 11.8% 4.5% 1.8% Backup 5,141 5,135 4,920 5,252 4,820 -17.7% -0.1% -4.2% 6.7% -8.2%

Source: Access Services

Figure 1: Monthly Trip Requests (FY2004 – FY2013)

175 200 225 250 275 300 325 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013

(13)

PASSENGER DEMAND FORECASTING

Trip demand is rising at a significantly faster pace in geographically smaller service regions than the other regions. Over the last ten fiscal years, trip demand in in Santa Clarita and Antelope Valley has increased rapidly. In 2009 and 2010, trip requests in Santa Clarita County nearly doubled, growing from 27,319 requests in 2008 to more than 48,584 in 2010. Since then trip demand in Santa Clarita has significantly slowed down. In Antelope Valley, trip requests grew by 25 percent per year on average over the past five fiscal years. In particular, demand increased by 33.5 percent in FY2013, the largest growth experienced by any service region. The growth in Antelope Valley accounted for 15 percent of total trip growth, although less than three percent of all trip requests came from Antelope Valley in 2012.

Figure 2 depicts the seasonality of paratransit demand, attributed in part to changing weather conditions, over the past five fiscal years. There is a common pattern in variations of trip demand over a twelve-month period. Trip requests tend to peak in spring and October; during summer and winter twelve-months the figures are lower in comparison (December, January, and February are the rainiest months in Los Angeles).

Figure 2: Seasonality of Paratransit Demand (FY2009 – FY2013)

Passenger Trips Completed

Although trip requests represent the fundamental manifestation of consumer demand, not all requested trips are scheduled. Requests can be denied by the agency because of eligibility requirements or cancelled by the customer after seeing the expected pickup time. After a trip is scheduled, the agency sends a vehicle to the pickup location. However, not all of these scheduled trips are completed due to customer no-shows and late cancellations. A paratransit agency incurs costs on trips that are scheduled and not completed whereas completed trips generate revenue for the agency.

200 230 260 290 320 350

Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun

FY 2009 FY 2010 FY 2011 FY 2012 FY 2013

Seasonal Trends: - Increase in October - Decline in Winter - Rebound in March

(14)

The number of passenger trips completed is the “realized” part of paratransit demand. Passenger trips completed can be divided into six categories: certification trips, ambulatory passengers, wheelchair passengers, personal care attendants (PCA), companions and children five years old and under.

Ridership

The number of trips completed, or ridership, is closely related to the number of trip requests and both have experienced similar trends over the past decade. As evidenced by the trend line in Figure 3, ridership remained stagnant from FY2003 to FY2007. Ridership increased in all regions in FY2008 and FY2009 before decreasing in FY2010 as a result of indirect effects of the economic recession. Since 2010, ridership has steadily increased and reached its highest level in FY2013. Over the last three years, ridership has grown by an average annual rate of 7.9 percent in the service area.

Figure 3: Ridership in Service Area (FY2004 – FY2013)

As shown in Table 2 on the next page, ridership has increased in every service region since FY2010. Of all the regions, Antelope Valley has demonstrated the strongest growth in ridership in the past decade. Over the previous four years, ridership in Antelope Valley nearly tripled, rising from 44,000 in FY2009 to 116,000 in FY2013. In FY2013 alone, ridership in Antelope Valley increased by 33 percent.

Smaller regions like Antelope Valley and Santa Clarita have had the largest growth rates of all regions, but most of the new ridership in the previous three years has come from the steady growth of the largest regions in Los Angeles County (Northern, Eastern, Southern, West/ Central). The Southern region has had the most ridership of all regions since FY2008. In years following the economic recession in FY2010, ridership increased by over 200,000 in the Southern region, which is the most of any other region in Los Angeles County. The Eastern region is the second largest in the county and has experienced modest growth since the recession, averaging an annual increase of 4.7 percent. Ridership in the Northern region has grown steadily and is one of the few regions to experience positive growth every year since FY2007.

150 175 200 225 250 275 300 325 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013

(15)

PASSENGER DEMAND FORECASTING

Table 2: Ridership by Service Region (FY2009 – FY2013)

FY2009 FY2010 FY2011 FY2012 FY2013

TOTAL 2,812,307 2,768,346 2,983,849 3,275,021 3,481,204 10.1% -1.6% 7.8% 9.8% 6.3% Antelope Valley 42,177 68,086 73,818 82,583 111,263 19.8% 61.4% 8.4% 11.9% 34.7% Eastern 846,470 844,727 887,614 938,910 977,840 7.4% -0.2% 5.1% 5.8% 4.1% Northern 525,709 532,403 582,464 648,509 668,668 10.7% 1.3% 9.4% 11.3% 3.1% Santa Clarita 28,392 38,939 44,372 44,622 46,381 26.0% 37.1% 14.0% 0.6% 3.9% Southern 907,832 843,838 898,519 1,029,309 1,131,881 16.6% -7.0% 6.5% 14.6% 10.0% West/ Central 457,241 435,818 492,801 526,465 540,810 1.9% -4.7% 13.1% 6.8% 2.7% Backup 4,486 4,535 4,261 4,623 4,361 -19.4% 1.1% -6.0% 8.5% -5.7%

Figure 4 on the next page shows the ridership trend for the all service regions in the past ten fiscal years. Several regions experienced decreases in ridership due to changes in service boundaries and in

contractors. Changes in West/ Central and Southern regional boundaries in FY2007 and again in FY2008 are evidenced by the drastic changes in ridership in those years. After the boundaries change, West/ Central ridership fell by ten percent in FY2007 and again by 17 percent in FY2008. During the same period, ridership in the Southern region increased by three percent and then by 21 percent.

(16)

PASSENGER DEMAND FORECASTING

Figure 4: Ridership in the Four LA Basin Regions (FY2004 – FY2013)

Distribution of Ridership by Service Region

During FY2013, the Southern region had the largest ridership share of 32.6 percent followed by the Eastern region at 28.1 percent. The Northern and West/ Central regions accounted for 19.2 percent and 15.6 percent of ridership respectively. The Santa Clarita and Antelope Valley regions accounted for less than five percent of total ridership combined. Figure 5 displays the distribution of passengers by service region in FY2009 through FY2013.

Despite the positive growth of the Eastern and West/ Central regions over the previous three years, their share of ridership decreased with respect to the other regions in Los Angeles County. Since FY2011, the Southern region increased its share of ridership by 2.4 percent and Antelope Valley by 0.7 percent, the most of any other region during that span. Otherwise, the ridership shares of the Northern and Santa Clarita regions have remained about the same.

0 20 40 60 80 100 120 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 Th o u san d s

(17)

PASSENGER DEMAND FORECASTING

Figure 5: Distribution of Passengers by Service Region (FY2009 – FY2013)

Ridership by Type of Passenger

Trips completed by the agency can be divided into six categories: certification trips, ambulatory

passengers, wheelchair passengers, personal care attendants (PCA), companions and children five years old and under.

Ambulatory passengers have consistently been the most served over the past decade by paratransit. In FY2013, ambulatory passengers accounted for 58.1 percent of total ridership. The majority of the remaining trips have been taken by persons using wheelchairs (17.8 percent) and PCA (21.6 percent). Over the years, trips completed by persons using wheelchairs have decreased while those by PCA have increased. The rest of passenger trips were distributed among companions, children, and certification trips. These passengers had a share of less than two percent each of total completed trips. Figure 6 depicts the distribution of ridership by type of passenger over the last ten fiscal years.

30.1% 30.6% 29.8% 28.7% 28.1% 16.3% 15.8% 16.5% 16.1% 15.6% 18.7% 19.3% 19.5% 19.8% 19.2% 32.3% 30.5% 30.2% 31.5% 32.6% 1.5% 2.5% 2.5% 2.5% 3.2% 1.0% 1.5% 1.5% 1.4% 1.3% FY 2009 FY2010 FY2011 FY2012 FY2013

(18)

PASSENGER DEMAND FORECASTING

Figure 6: Ridership by Type of Passenger (FY2004 – FY2013)

Key Operating Factors

Demand for paratransit service is affected by multiple factors, including fare structure, operating

standards, and other socioeconomic indicators. Key operating factors that are considered to be impacting trip demand can include inflation adjusted trip fare (real fare), complaint rate, on-time performance, cancellation and no-shows, and population.

Real Fare

Economic theory and past experience show an inverse relationship between real fare (as opposed to nominal fare) and trip requests, sometimes with a slight time lag: a reduction in real fare typically

generates an increase in trip requests, whereas an increase in real fare typically generates a reduction in trip requests. The extent of this relationship is measured by the elasticity of demand with respect to real fare, which measures demand responsiveness with respect to price.

The average real fare is computed for each region in two steps. First, the average nominal fare is computed by dividing total monthly fare revenue (cash, Access Services coupons and Metropolitan Transit Authority (MTA) bus tokens) by the number of passengers who paid for the trip (i.e., ambulatory riders, wheelchair users, and companions). Personal care attendants, and children five years old and under (if traveling as companions) do not pay the fare, as well as other passengers on certification trips. Next, the average nominal fare is deflated by the Consumer Price Index (CPI) for Los Angeles-Riverside-Orange County, CA. This removes all inflationary movements from the nominal fare price, allowing the fare to be expressed in constant dollars (or 2013 dollars). Figure 7 shows the trend in regional average real fare along with the number of trip requests since July 2001.

59.3% 58.6% 57.1% 56.4% 56.9% 57.6% 56.0% 57.0% 57.4% 58.1% 20.2% 20.7% 21.0% 21.0% 20.2% 19.0% 18.7% 17.7% 17.7% 17.8% 1.4% 1.5% 1.7% 2.0% 1.9% 1.8% 1.6% 1.5% 1.5% 1.6% 1.8% 1.4% 1.7% 1.9% 1.9% 1.5% 1.6% 1.8% 1.8% 1.8% 16.8% 17.6% 18.0% 18.2% 18.7% 19.7% 21.8% 21.5% 21.5% 21.6% 0.5% 0.3% 0.5% 0.4% 0.4% 0.4% 0.3% 0.3% 0.4% 0.4% FY 2004 FY 2005 FY 2006 FY 2007 FY 2008 FY 2009 FY 2010 FY 2011 FY 2012 FY 2013

(19)

PASSENGER DEMAND FORECASTING

Figure 7: Trip Requests and Average Real Fare (FY2004 – FY2013)

Sources: Access Services and California Department of Finance

From the figure above, the following noteworthy points stand out:  The average real fare follows a downward trend;

 The fare change in July 20061 induced little, if any, volatility in trip fare. The monthly average nominal fare up until the last fare change in July 2009 remained stable, trending within a narrow range between $1.88 and $1.90; and

 The last change in fare structure that occurred in July 2009 led to an increase in the average nominal fare from $1.89 in June 2009 to $2.32 the following month.

Throughout the study period, changes in nominal fare have induced changes in trip demand. The start of the 2007 fiscal year was met with a drop in trip requests from 221,926 in June of 2006 to 215,352 in July. The 2009 fare increase led to a reduction in ridership from 252,253 in June to 246,582 in July, followed by another decline of 13,379 in trips taken in a month later.

Eligibility Evaluations and New Applicants

Eligibility evaluations consist of evaluations of individual as new applicant or recertification for paratransit services. For Access, eligibility is determined by an in-person transit evaluation and is based on the individual’s ability to use accessible buses and trains in LA County. Evaluation is not based solely on the disability, age, or medical conditions.

The discussion of eligibility evaluations and new applicant is provided in Chapter 7, along with the analysis of new applicant to investigate the rapid growth in new applicants since 2009.

1

The new (reduced) fare for trips scheduled between 9:00 p.m. and 5:00 a.m. is $1.50 regardless of distance. $1.50 $1.75 $2.00 $2.25 $2.50 $2.75 $3.00 $3.25 0 50 100 150 200 250 300 350 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012

Trip Requests (Thousands) Real Average Fare (2013 Dollars, Right Axis)

Change in fare structure

(20)

Complaint Rate

The complaint rate, defined as the number of passenger complaints per one thousand passengers carried, reflects the quality of the service received by customers. Since demand is partly defined by the willingness to pay, it is expected that decreases in the complaint rate result in increases in the number of trip requests (typically with a lag of one or more months) and vice versa. This is depicted in Figure 8 where the data in the past fiscal years show that improvements in the quality of service coincide with lagged increases in ridership (and vice versa). However, since complaint rate is a function of service provided, it and not an independent variable and it is not included in the analysis.

Figure 8: Ridership and Complaint Rate (FY2004 – FY2013)

Figure 9 on the following page reports the data by service region.

0 2 4 6 8 10 12 14 16 18 20 0 50 100 150 200 250 300 350 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013

Trip Requests Complaint Rate (Right Axis) Compliants (Per Thousand) Trip Requests (Thousands)

(21)

PASSENGER DEMAND FORECASTING

Figure 9: Ridership and Complaint Rate by Service Region (FY2009 – FY2013)

0 20 40 60 80 0 20 40 60 80 2008 2009 2010 2011 2012 2013

Trip Requests (Thousands) Complaints Per Thousand Trips

West/ Central 0 3 6 9 12 0 30 60 90 120 2008 2009 2010 2011 2012 2013

Southern 0 2 4 6 8 0 20 40 60 80 2008 2009 2010 2011 2012 2013

Northern 0 2 4 6 8 10 0 20 40 60 80 100 2008 2009 2010 2011 2012 2013

Eastern 0 2 4 6 8 10 12 14 0 2 4 6 8 10 12 14 2008 2009 2010 2011 2012 2013

Antelope Valley

One-time adjustment for previously under-reported PCAs and children

0 1 2 3 4 5 6 0 1 2 3 4 5 6 2008 2009 2010 2011 2012 2013

(22)

On-Time Performance

At the system level, on-time performance averaged 90.4 percent over the period FY2003 – FY2013. This estimate is above the 90 percent benchmark set by Access in the Year 2000 Strategic and Short-Term Business Plan. As shown in Figure 10 below the upward trend in on-time performance in FY2013 was interrupted several times with significant drops below the benchmark. These declines coincide with lagged decreases in trip requests, possibly a result of the implementation and/or suspension of reservation, scheduling, and dispatching software modules.

Figure 10: Trip Requests and On-Time Performance (FY2004 – FY2013)

Cancellations and No-shows

The no-show rate is defined as the number of no-shows divided by the number of trip requests. Likewise, the cancellation rate is defined as the number of cancellations divided by the number of trip requests. Historically, these two measures shared similar trends throughout the past ten fiscal years: rising during the first five and then dropping significantly from FY2007 due to a policy change2. In FY2013, the no-show rate was estimated at 1.6 percent only – compared to 4.6 percent in FY2004. Overall, fluctuations in no-show and cancellation rates do not coincide with changes in trip requests as illustrated in Figure 11 on the next page.

2

A late standing order cancellation policy has been effective since February 1st, 2007. Under this new policy, riders are allowed an unlimited number of cancellations, as long as they are made by 10:00 p.m. the night before service. Trips that are cancelled after this time are classified as late standing order cancellations. A rider is allowed a maximum of six late standing order cancellations (or 10 percent of his/her trips, whichever is greater) in a 60-day period. Riders who cancel more often than this are subject to revocation of their standing order trip.

80% 84% 88% 92% 96% 100% 0 50 100 150 200 250 300 350 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013

(23)

PASSENGER DEMAND FORECASTING

Figure 11: Trip Requests and No-Show and Cancellation Rates (FY2004 – FY2013)

Population

The demand for paratransit services may also be affected by the number of people living in the service area. Population data for Los Angeles County are collected from the Demographic Research Unit at the California Department of Finance (DoF). An average annual growth rate of 0.66 is estimated between 2013 and 2014; from then on, rates are expected to drop to 0.62 percent.

On the other hand senior population, which is defined by the DoF as 85 years old and above, is growing at much faster pace. In particular, between now to 2014, this group of population will grow at over three percent and continue at a two percent rate afterwards and it is due to the growing Asian and Latino Americans senior population. Figure 12 and Figure 13 on the following page illustrate the trends in population. 0% 1% 2% 3% 4% 5% 6% 7% 8% 9% 0 50 100 150 200 250 300 350 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013

(24)

PASSENGER DEMAND FORECASTING

Figure 12: Population Growth in Los Angeles County (2003 – 2018)

Source: California Department of Finance, Demographic Research Unit

Figure 13: Senior Population Growth in Los Angeles County (2003 – 2018)

Source: California Department of Finance, Demographic Research Unit 8.5 8.7 8.9 9.1 9.3 9.5 9.7 9.9 10.1 10.3 10.5 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 M illi o n s Forecast Historical 0 20 40 60 80 100 120 140 160 180 200 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 Tho u san d s Forecast Historical

(25)

PASSENGER DEMAND FORECASTING

3.

Performance Metrics

Adding to the review of Access’s operations and performance statistics is a comparative assessment of key performance metrics from a sample of paratransit agencies. The comparison sheds some light on how performance is being tracked and monitored by different agencies and the assessment may help Access to develop initiatives for establishing new performance goals in the future. Furthermore, ongoing oversight of performance can help Access plan for the lingering impact of the economic recession in terms of unemployment and tax revenue (primary funding source for public transit), as well as uncertainty in gasoline prices. The discussion also serves as an introduction to the peer analysis of Access’s

operations and quality of service, which is presented in Section 4.

Agencies establish and track performance metrics for reporting, planning, and funding purposes. In this section, a set of key performance metrics additional to those already introduced in earlier sections are presented for a sample of paratransit service systems. While details of the selection process of the comparison agencies are explained in the peer analysis (Section 4), agencies are selected primarily because the metrics are recorded in agency reports that are readily available online. Metrics that are measured differently from those introduced in earlier sections are provided with definitions or

explanations on how they differ from those provided by Access.

Overall, Access has been reporting similar metrics in terms of service delivery and coverage. To maintain transparency and accountability, Access may consider providing in its annual report additional metrics on service solvency, completeness, and maintenance such as subsidy per passenger, vehicle no-shows (“missed trips”), and miles between road calls, etc. In terms of safety, Access may consider reporting total accidents that aggregate the numbers provided in the different management summaries of the Board Box Report.

Other findings are as follows:

 Customer complaint rates are usually measured by the number of complaints per 1,000 trips. For Pace Suburban Bus Division (PACE), the metric is measure by complaints per 100,000

passenger miles;

 Washington Metropolitan Area Transit Authority (WMATA), Orange County Transportation Authority’s (OCTA), and Access measure excessively late vehicles slightly differently. WMATA reports any trips over 30 minutes past window; Access reports “late 4” trips – category of late trips wherein the vehicle arrives more than 45 minutes after the end of the 20-minute on-time window; while OCTA measures service delivery failures (SDF), a unique measurement specific to the program. This indicator is an occurrence when a vehicle does not arrive at the pick-up location until 90 minutes after the conclusion of a 30-minute on-time window;

 All sample agencies publish accidents rates except for WMATA. Preventable vehicle accidents are counts of incidents concerning physical contact between a paratransit vehicle and other vehicles, objects, or pedestrians where the operator is determined to be at fault. The standardized

measurement is accident counts multiplied by 100,000 and then divided by the total vehicle miles;  PACE and OCTA publish miles between road calls, a maintenance performance indicator that

measures the vehicle miles between mechanical failures of a vehicle used for public transit during revenue service. Road calls may cause a delay in service and necessitate removing the vehicle from service until repairs are made; and

 Subsidy per passenger is reported by PACE. Subsidy includes Public Transportation Fund (PTF) of 30 percent of the Regional Transportation Authority (RTA) sales tax and Chicago real estate transfer tax (RETT) collected.

(26)

PASSENGER DEMAND FORECASTING

Table 3: Sample of Performance Metrics Published in Annual/ Monthly Reports

Metrics Access OCTA WMATA PACE MDT

Service Coverage

Total Passengers     

Total Trips Requested  

Total Trips Scheduled

Total Trips Delivered   

Contract Revenue Miles 

Contract Revenue Hours 

Average Trip Distance 

Vehicles in Service 

Passengers per Hour  

Service Delivery

On-Time Performance     

Hour Late Trips  

Service Complaints    No-Show (Customer)      No-Show (Vehicle)   Late Cancelation    Service Solvency

Cost per Revenue Vehicle Hour   

Subsidy per Passenger 

Farebox Recovery Ratio*  

Service Safety Preventable Vehicle Accidents    

Other Miles between Road Calls  

Average Initial Hold Times or Call Response**  

Sources: OCTA −Transit Division Performance Measurements Report; WMATA − MetroAccessMonthly Operations Report; PACE − Suburban Service Budget &Regional ADA Paratransit Budget; MDT − Miami-Dade County Transit (Miami, FL) Paratransit Operations Monthly Report.

Notes: *Farebox recovery ratio is a measure of the proportion of operating costs covered by passenger fares; calculated by dividing the farebox revenue by total operating expenses. **Metrics refers to customer service delay in seconds.

(27)

PASSENGER DEMAND FORECASTING

4.

Peer Analysis

A peer review is a valuable management tool designed to help improve an agency’s service and operation performance. Ultimately, the goal of the peer review is to better understand an agency’s strengths and weaknesses so as to formulate strategies to improve its performance. For Access, the objective of the customized peer review is to compare similar paratransit agencies (in terms of operational statistics, size, and geography) to identify demand-related issues (such as increased customer

complaints, high no-show rate and transfer of ridership from other specialized service providers) that have risen elsewhere and to examine how these issues were addressed. The findings may also be useful to Access management in formulating policy scenarios.

Methodology

The peer review approach relies on a methodology developed for the Transportation Research Board (TRB)3, which consists of the following five steps:

1. Define the performance areas to be assessed;

2. Establish a peer group based on guidance provided by Access and using the FTIS database; 3. Gather and process performance data for all selected peers; and

4. Compare performance data and identify areas of improvement.

The selection of peer group is primarily based on operational statistics, size and geography, as well as the study team’s prior experience with different transit agencies in obtaining relevant data. To verify that the appropriate agencies are selected, likeness scores computed within the FTIS database are utilized4. The resulting six agencies form the national peer group:

 Massachusetts Bay Transportation Authority (Boston, MA);  Metropolitan Transit Authority of Harris County (Houston, TX);  Miami-Dade Transit (Miami, FL);

 Pace Suburban Bus Division (Chicago, IL);

 Southeastern Pennsylvania Transportation Authority (Philadelphia, PA); and  Washington Metropolitan Area Transit Authority (Washington, D.C.).

To assess how Access performs within the Los Angeles region, a group of regional peer agencies are selected based on relative proximity to the region. The four selected agencies are:

 Orange County Transportation Authority (OCTA);  Riverside Transit Agency (RTA);

 LACMTA - Small Operators (LACMTA); and

 City of Los Angeles Department of Transportation (LADOT).

A standard peer review requires a level of effort that exceeds the current scope of the demand forecasting. Instead, a selection of performance areas of interest to Access is assessed. More specifically, the following five areas have been considered:

3

Kittelson & Associates, Inc. et al. A Methodology for Performance Measurement and Peer Comparison in the Public

Transportation Industry. TCRP Report 141, Transportation Research Board, National Research Council, Washington,

D.C., 2010.

4

The scores determine the level of similarity between a potential peer agency and the target agency with respect to a number of screening/grouping criteria accounting for both an agency’s operating characteristics (annual vehicle miles operated, annual operating budget, etc.) and the socio-economic profile of the service area (population, percentage of low-income people, etc.). A total likeness score is then calculated. A total likeness score of 0 indicates a perfect match between two agencies. Higher scores denote greater levels of dissimilarity between two agencies. In general, a total likeness score lower than 0.50 indicates a good match, a score between 0.50 and 0.74 represents a

satisfactory match, and a score between 0.75 and 0.99 suggests that potential peers may be available, but caution should be exercised to investigate potential differences that may make them unsuitable. Finally, peers with scores greater than or equal to 1.00 should not be considered in a performance peer review.

(28)

PASSENGER DEMAND FORECASTING

 Service utilization – measures how passengers use the service that is provided5

: Passenger trip is the demand of service and it is the main indicator of service utilization. Passenger trip is also used to compute two other important indicators: 1) Late cancellation rate, which is the percentage of trips cancelled less than two hours within the negotiated time window and, 2) No show rat, which is the percentage of trips where customers did not show up within the allotted 20-minute pick up time window or cancel a Standing Order6 trip later than 10pm of the day prior to schedule pick-up;

 Cost efficiency – assess an agency’s ability to provide service outputs within the

constraints of service inputs7: Operating cost per passenger trip is the cost to provide service for each passenger demanding the service and it compares the cost of providing service to the outcomes (trips made) resulting from the provided service. Cost components included in

operating cost are wages and fringe benefits, utilities, causalities and liabilities, services, fuel and lube, tire, etc.;

 Productivity – look at how many passengers are served per unit of service—hours, miles, vehicles, or employee full-time equivalents8: Passenger per revenue hour compares demand of the provided service to a time-specific unit of service;

 Cost effectiveness – compare the cost of providing service to the outcomes resulting from the provided service9: Farebox recovery ratio measures how much of a transit agency’s

operating costs are covered by fare revenue and the agency’s ability to recover (in full or in part) the cost of providing transit service. Revenue generated is used as the outcome resulting from the provided service;

 Service Quality (Perceived) – describe the transit agency’s service as perceived by

customers: On-time performance demonstrates the level of satisfaction that passenger of the service experience. A trip is considered on time if vehicle arrives within a 20-30 minute pick-up window.

The review of the areas of interest introduced covers data from FY2008 to FY2011 to account for short-term trends and identify potential outliers in the data during the four-year period. The data are collected from the following sources:

 Florida Transit Information System (FTIS)10;  National Transit Database (NTD);

 New York City Transit Paratransit Peer Reports and  Agency Operation and Service Annual Reports1112.

Note also that all monetary metrics are adjusted for inflation and expressed in constant 2013 dollars using the U.S. Consumer Price Index (CPI). Removing inflation allows a trend analysis to clearly show whether an agency’s real costs are increasing or decreasing.

5

Transit Cooperative Research Program Report 141: A Methodology for Performance Measurement and Peer Comparison in the Public Transportation Industry. 2010.

6

A Standing Order trip is a series of pre-scheduled trips based on repeated trips of same time and destinations, for an extended period of time on the same day(s) of the week.

7

Transit Cooperative Research Program Report 141: A Methodology for Performance Measurement and Peer Comparison in the Public Transportation Industry. 2010.

8

Ibid.

9

Ibid.

10

Available at: http://www.ftis.org/

11_{OCTA −Transit Division Performance Measurements Report; WMATA − MetroAccess Monthly Operations Report;}

PACE − Suburban Service Budget & Regional ADA Paratransit Budget; MDT − Miami-Dade County Transit (Miami, FL) Paratransit Operations Monthly Report.

12

Data on service quality are somewhat incomplete, especially for complaint rate and late cancellation and those data are not readily available.

(29)

PASSENGER DEMAND FORECASTING

Service Utilization

Passenger demand in terms of passenger trips or trip requests is an indicator of service utilization. Because the number of passenger trips is commonly reported and provided by each agency, it is used to quantify demand in the peer review. Other demand measures, such as trip requests, are less readily available. Access is one of the largest paratransit agencies in terms of passenger demand, providing the third most number of trips amongst all paratransit systems nationwide in 2011. The only systems larger than Access in terms of ridership were PACE in Chicago, which experienced a drastic 30 percent increase in ridership in 2011, and Metropolitan Transportation Authority (MTA) New York City Transit. Ridership for the selected peer systems are displayed in Figure 14.

Access averages about 2.8 million passenger trips a year which is well above the median value of 1.6 million for national peer systems. The average annual ridership growth for Access is 5.5 percent, which is the fourth highest amongst national peers behind agencies in Chicago, Washington and Boston.

Figure 14: National Peer Review, Number of Passenger Trips (FY2008 – FY2011)

Access is the largest and the fastest growing paratransit agency in Greater Los Angeles region. Ridership for some other agencies in the region is declining while Access continues to increase its number of passenger trips each year. On average, Access serves nearly twice as many passengers as OCTA, which is the next largest paratransit agency in Los Angeles. The median value for ridership amongst regional peers is 1.2 million which is substantially lower than the Access average of 2.8 million passengers per year. The number of passenger trips for other agencies in the region are displayed in Figure 15. 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0

Boston Houston Miami Chicago Philadelphia Washington Access

M ill io n s 2008 2009 2010 2011

(30)

PASSENGER DEMAND FORECASTING

Figure 15: Regional Peer Review, Number of Passenger Trips (FY2008 – FY2011)

The economic crisis may have affected ridership numbers in the Greater Los Angeles region in 2010. The recession led to a decline in tax revenues which translated into funding shortages for paratransit agencies in the area. Many agencies responded by cutting service, revising policies and increasing fares which resulted in ridership decreases. Every agency experienced ridership decrease in ridership in 2010 and the fall in ridership is most evident in LADOT and LACMTA in 2010.

No show and late cancellation rates are also considered drivers of service utilization as they indicate the percentage of trips that were scheduled, but not completed. These are important to include because agencies incur costs but do not generate revenue on these trips. The no show and late cancellation rates for the national peer systems are displayed in Figure 16 and Figure 17, respectively.

Figure 16: National Peer Systems, No Show Rates (FY2008 – FY2011)

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0

LADOT LACMTA Orange County Riverside Access

M

ill

ion

s

2008 2009 2010 2011

Median Value = 1.2 Million

0% 1% 2% 3% 4% 5% 6% 7% 8%

Boston Houston Miami Chicago Philadelphia Washington Access 2008 2009 2010 2011

(31)

Figure 17: National Peer Systems, Late Cancellation Rates (FY2008 – FY2011)

Amongst its national peer systems, Access has the lowest rate of passenger no shows and the third lowest rate of late cancellations. The average no show rate for Access is 1.79 percent which is below the median value of 2.60 percent. The average rate of late cancellations is 2.66 percent which is also below the median value of 2.96 percent. The data indicate that Access Services has a lower percentage of scheduled trips that do not generate revenue in comparison to other national peer systems.

Cost Efficiency

System cost efficiency is quantified as the operating cost per passenger trip. Operating costs include the total expenses to operate and maintain the transit system, which includes labor, fuel, maintenance, taxes and other costs associated with transit operations. According to the NTD 2011 profiles for top 50 reporter agencies, employee benefits and wages typically account for at least half of all operations and

maintenance expenses. The operating cost per trip is expressed in 2013 dollars and is displayed for all national and regional peer systems in Figure 18 and Figure 19.

The average operating cost for Access from the data is about equal to the median value for national peer systems of $34.3. There is a disparity in operating costs for other national paratransit systems. Paratransit agencies in Chicago, Washington, and Boston all have average operating costs over 40 dollars per trip while agencies in Philadelphia, Houston and Miami have average costs under 30 dollars per trip. This could be due to a discrepancy in employee compensation across systems. The real cost per trip for Access remains relatively steady from 2008 to 2011, growing 1.2 percent per year on average.

0% 4% 8% 12% 16% 20%

(32)

PASSENGER DEMAND FORECASTING

Figure 18: National Peer Review, Real Operating Cost per Passenger Trip (FY2008

– FY2011)

Though comparing favorably to national peer systems, the operating cost per trip for Access Services is higher than the median value of 28 dollars per trip for other systems in the Los Angeles region (as shown in Figure 19). In terms of operating cost per trip, Access does not seem to benefit from economies of scale by having more rider than the other regional peer agencies. This could be because Access covers a larger service area that spans multiple regions, increasing the length of each trip.

Figure 19: Regional Peer Review, Operating Cost per Passenger Trip (FY2008 –

FY2011)

Regional average trip length may be important to consider when assessing the cost efficiency of an agency. An agency covering a large service area such as Access may be at a disadvantage in terms of cost efficiency because vehicles have to cover longer distance to deliver services, thus making trips more expensive to provide. In 2008, Access had the longest average trip length in the Los Angeles region and the second largest average length amongst national peer systems with an average of 12.7 miles traveled per trip. Only Miami had a longer average trip length at 13.6 miles. The trip length for Access is more than three times longer than that of LADOT and LACMTA, which are 3.9 and 3.7 miles, respectively. Riverside Transit Agency had a comparable average trip length to Access of 12.3 miles per trip.

$0 $10 $20 $30 $40 $50 $60

Boston Houston Miami Chicago Philadelphia Washington Access 2008 2009 2010 2011 Median Value = $34.3 $0 $10 $20 $30 $40 $50 $60

LADOT LACMTA Orange County Riverside Access 2008 2009 2010 2011

(33)

PASSENGER DEMAND FORECASTING

Productivity

The number of passengers per revenue hour indicates how many passengers an agency serves for each hour that vehicles are earning revenue. Agencies that serve more passengers per hour are deemed more productive. The number of passengers per revenue hour for Access and all national peer systems is shown in Figure 20.

Figure 20: National Peer Review, Passengers per Revenue Hour (FY2008 –

FY2011)

Access averages 1.8 passengers per revenue hour which was higher than all of the national peer

systems except SEPTA in Philadelphia which averages 1.8 passengers. The median value for all national peer systems is 1.7 passengers. The least productive system is WMATA which averages 1.1 passengers per revenue hour.

The number of passengers per revenue hour is a good indicator of productivity in a system but has some drawbacks as a metric. The size of the service area and trip length can greatly affect the number of passengers per revenue hour. Access has the highest service area population of national peer systems, covering nearly 12 million people in the Los Angeles region.

Access covers a substantially larger service area than other national systems and is still one of the most productive in terms of passengers per revenue hour. However, Access has a lower rate of productivity than regional peer agencies. The median value amongst peer agencies in the Great Los Angeles region was 2.14 passengers per revenue hour as shown in Figure 21.

0.0 0.5 1.0 1.5 2.0 2.5

(34)

PASSENGER DEMAND FORECASTING

Figure 21: Regional Peer Review, Passengers per Revenue Hour (FY2008 –

FY2011)

The discrepancy in passengers per revenue hour amongst national and regional peer systems could be due to relative service areas of smaller regional agencies. By covering a smaller area, it is easier to serve more passengers per revenue hour because less time is spent traveling to pick up and deliver the

passenger to their destination.

Cost Effectiveness

Farebox recovery ratio is the percentage of passenger fare revenues out of total operating expenses. As discussed earlier, factors such as wages, benefits, fuel, insurance, maintenance and trip length all contribute to the operating cost for each paratransit agency. The farebox recovery ratio is an indicator of the share of total operating costs that is covered by passenger fares. It is used to quantify cost

effectiveness because it measures the return of each dollar as revenue over cost. A higher percentage means that passenger fares made up a greater portion of the agency’s operating costs.

0.0 1.0 2.0 3.0 4.0

(35)

PASSENGER DEMAND FORECASTING

Access Services has an average farebox recovery ratio of 5.20 percent, which is on par with the median value of other national peer systems of 5.24 percent. Farebox recovery ratios for national peer systems are displayed in Figure 22. The farebox recovery ratio for Access Services increases from 5.0 percent to 5.5 percent during the observation period, which was the largest increase amongst national systems except for Boston which increased from 2.9 percent to 4.1 percent during the same period.

Figure 22: National Peer Review, Farebox Recovery (FY2008 – FY2011)

The farebox recovery ratios for regional peer systems are displayed in Figure 23. The median value amongst regional peers is 5.5 percent, which is close to the average for Access. Access has a higher farebox recovery ratio than LADOT and LACMTA, but has a lower ratio than agencies in Orange County and Riverside.

Figure 23: Regional Peer Review, Farebox Recovery (FY2008 – FY2011)

Service Quality

Many agencies do not have data for metrics pertaining to service quality readily available, making a peer system comparison difficult. Of the metrics available, on time performance is the only indicator provided by nearly all of the peer agencies. Agencies differ slightly on their definition of the time window that constitutes a trip being completed on time, with pick-up window ranging from 20-30 minutes from the

0% 2% 4% 6% 8% 10% 12% 14% 16%

Boston Houston Miami Chicago Philadelphia Washington Access 2008 2009 2010 2011 Median Value = 5.24% 0% 2% 4% 6% 8% 10% 12% 14% 16%

(36)

PASSENGER DEMAND FORECASTING

schedule pick-up time. Figure 24 and Figure 25 show the percentage of trips that were completed on time amongst national and regional peer systems.

The percentage of trips completed on time for Access Service is on par with the median value for national agencies, averaging about 91 percent from 2008 to 2011. Compared to regional peer systems, on time performance for Access is worse than its regional peers in LACMTA and Orange County. On time performance for Access improves over the analysis period, increasing from 89.9 percent to 91.4 percent.

Figure 24: National Peer Review, On Time Performance (FY2008 – FY2011)

Overall, Access Services compares favorably with other national paratransit agencies, with most performance metrics grading out around the median level. Unlike some other agencies with measures that fluctuate over time, Access performs consistently in terms of service provided and perceived. The analysis of regional trends suggests that Access may consider investigating into reasons for other agencies being able to achieve higher passengers per revenue hour with lower operating cost per passenger trip.

Figure 25: Regional Peer Review, On Time Performance (FY2008 – FY2011)

70% 75% 80% 85% 90% 95% 100%

Boston Houston Miami Chicago Philadelphia Washington Access 2008 2009 2010 2011 Median Value = 91.54% 70% 75% 80% 85% 90% 95% 100%

(37)

Passenger Demand Forecasting

February 2014

PASSENGER DEMAND FORECASTING

CONTENTS

PASSENGER DEMAND FORECASTING

PASSENGER DEMAND FORECASTING

LIST OF TABLES

LIST OF FIGURES

PASSENGER DEMAND FORECASTING

PASSENGER DEMAND FORECASTING

Executive Summary

Demand Analysis

PASSENGER DEMAND FORECASTING

Table ES 1: Operations Forecasts, Mean Expected Outcome (FY2014 – FY2018)

Results 2014 2015 2016 2017 2018

Table ES 2: Ridership Forecasts by Service Region, Mean Expected Outcomes

Fiscal

Analysis of New Applicants

Table ES 3: New Applicant Forecasts by Service Region, Mean Expected

Fiscal

Introduction

PASSENGER DEMAND FORECASTING

Plan of the Report

PASSENGER DEMAND FORECASTING

Historical Overview

PASSENGER DEMAND FORECASTING

Table 1: Passenger Trip Requests by Region (FY2009 – FY2013)

PASSENGER DEMAND FORECASTING

Figure 2: Seasonality of Paratransit Demand (FY2009 – FY2013)

Seasonal Trends: - Increase in October - Decline in Winter - Rebound in March

Ridership

PASSENGER DEMAND FORECASTING

Table 2: Ridership by Service Region (FY2009 – FY2013)

PASSENGER DEMAND FORECASTING

Figure 4: Ridership in the Four LA Basin Regions (FY2004 – FY2013)

PASSENGER DEMAND FORECASTING

Figure 5: Distribution of Passengers by Service Region (FY2009 – FY2013)

PASSENGER DEMAND FORECASTING

Figure 6: Ridership by Type of Passenger (FY2004 – FY2013)

Real Fare

PASSENGER DEMAND FORECASTING

Figure 7: Trip Requests and Average Real Fare (FY2004 – FY2013)

Eligibility Evaluations and New Applicants

Change in fare structure

Complaint Rate

PASSENGER DEMAND FORECASTING

Figure 9: Ridership and Complaint Rate by Service Region (FY2009 – FY2013)

Antelope Valley

On-Time Performance

PASSENGER DEMAND FORECASTING

Figure 11: Trip Requests and No-Show and Cancellation Rates (FY2004 – FY2013)

PASSENGER DEMAND FORECASTING

Figure 12: Population Growth in Los Angeles County (2003 – 2018)

PASSENGER DEMAND FORECASTING

Performance Metrics

PASSENGER DEMAND FORECASTING

Table 3: Sample of Performance Metrics Published in Annual/ Monthly Reports

Metrics Access OCTA WMATA PACE MDT

PASSENGER DEMAND FORECASTING

Peer Analysis

PASSENGER DEMAND FORECASTING

PASSENGER DEMAND FORECASTING

Service Utilization

PASSENGER DEMAND FORECASTING

Figure 15: Regional Peer Review, Number of Passenger Trips (FY2008 – FY2011)

Median Value = 1.2 Million

Figure 17: National Peer Systems, Late Cancellation Rates (FY2008 – FY2011)

PASSENGER DEMAND FORECASTING

Figure 18: National Peer Review, Real Operating Cost per Passenger Trip (FY2008

PASSENGER DEMAND FORECASTING

Productivity

PASSENGER DEMAND FORECASTING

Figure 21: Regional Peer Review, Passengers per Revenue Hour (FY2008 –

PASSENGER DEMAND FORECASTING

Figure 22: National Peer Review, Farebox Recovery (FY2008 – FY2011)

PASSENGER DEMAND FORECASTING

Figure 24: National Peer Review, On Time Performance (FY2008 – FY2011)