The PA Pulse is a timely and multimodal indicator created by the Port Authority to track the rise and fall of freight movement and passenger travel activity levels at the region's ports, airports, and interstate crossings. The index seasonally adjusts, weights, and combines raw transportation activity data to generate a measure that controls for predictable seasonal and calendar effects and short-term changes in travel patterns, enabling meaningful comparisons from month to month. The PA Pulse represents a broad cross section of transportation activity in the New York-New Jersey region, as the networks served by the Port Authority include many of the region's principal gateways and corridors for interstate commerce.
|Change vs. Previous
|Change vs. Previous
|PA Freight Pulse||94.4||-0.3%||1.1%|
|PA Passenger Pulse||101.1||2.9%||3.2%|
Index is Seasonally Adjusted, 2010=100
Last Updated: July 2014
In May 2014, the PA Pulse up 1.4 percent from its April 2014 value, and now stands about 2.2 percent higher than it did a year ago. The PA Freight Pulse went down by a seasonally adjusted 0.3 percent from last month, and the PA Passenger Pulse is up 2.9 percent from its April value.
|PA Pulse||PA Freight Pulse||PA Passenger Pulse|
The following table shows how the values of the indices have varied over the past year.
|PA Combined Pulse||PA Freight Pulse||PA Passenger Pulse|
|Month||Index||Pct. Change||Index||Pct. Change||Index||Pct. Change|
What is the PA Pulse?
The PA Pulse is a timely, multimodal, and seasonally adjusted indicator that captures freight and passenger activity levels at the region’s ports, airports, and interstate crossings. It represents a broad cross section of transportation activity in the New York-New Jersey region, as the facilities and networks served by the Port Authority are some of the principal gateways to the city and region. While the indices do not precisely account for every passenger or every dollar worth of freight that passes through Port Authority facilities, they approximate those quantities closely using observed data.
The PA Pulse tracks the rise and fall of regional transportation activity over time, controlling for predictable factors such as seasonal change and calendar effects (the number of weekdays and weekend days in each month, moving holidays, and leap years). The remaining variability in the index results from regional growth, changes in the regional economy, weather, and countless other factors that influence travel on the regional, national, and international transportation networks. The advantage of an index of this kind is that it provides an insight into whether regional transportation activity has increased or decreased on a month-to-month basis by accounting for recurring patterns of climate and calendar without unduly filtering out the remaining variations.
The PA Pulse is not designed to be a leading or coincident economic indicator; it does not represent or predict regional economic activity in any definitive way. While the index does show a correlation with trends in the regional and national economies, it is not a substitute for a carefully developed economic index such as the Federal Reserve Bank’s regional Index of Coincident Economic Indicators. Over time, we intend to pursue additional analysis in order to test the relationship between the PA Pulse and various regional economic indicators.
Finally, we emphasize that the PA Pulse is expected to rise and fall as regional conditions change, and is not a metric of agency, business unit, or system performance. The index is primarily driven by external changes in demand. Supply-side decisions about levels of service and management practices influence activity over time, but this is not measured directly by this index.
What are the components of the index?
The PA Pulse consists of two equally weighted components, the PA Freight Pulse and the PA Passenger Pulse. In turn, each is composed of variables that have been chosen to provide broad-based coverage of activity at key transportation gateways and on the region’s interstate transportation network.
The PA Freight Pulse combines four variables to measure goods movement at the Port Authority’s airports, port facilities, and bridges and tunnels:
The PA Passenger Pulse combines six variables to measure activity at the region’s airports and interstate crossings. Specific facilities and transportation services included in the PA Passenger Pulse include:
Variables not currently included in the index include Amtrak and general aviation passengers, as well as domestic waterborne freight. The index also excludes highway, transit, and rail activity that does not pass between New York and New Jersey.
Why was the PA Pulse created?
The PA Pulse represents part of the Port Authority’s effort to make its transportation data more accessible and useful. In 2011, the Port Authority began reporting transportation data in its Monthly Economic Indicators report and subsequently developed the PA Pulse to provide a comprehensive picture of activity on the agency’s regional transportation system. Traffic volumes and other transportation activity indicators do not facilitate easy analysis of trends due to seasonal variations and mode shifts that occur as a result of changing costs and service levels; thus, a seasonally adjusted index that incorporates a broad cross section of modes and facilities provides a more useful glimpse of regional transportation activity levels and their changes over time.
How can the PA Pulse be used?
The PA Pulse simply describes past transportation activity in a general manner and does not predict future transportation activity. The PA Pulse is an experimental indicator that offers a broad overview of regional transportation trends, and this descriptive function is its best application at present. We do not intend to use the PA Pulse to inform policy decisions, nor should external parties use it to draw conclusions about specific Port Authority operations or facilities. The number of factors that influence the PA Pulse is large - the index should not be interpreted as a simple measurement of transportation facilities’ performance.
How is the index calculated?
The PA Pulse combines four freight variables and six passenger variables, all of which are based upon monthly data on activity levels at Port Authority facilities (see above). All variables undergo seasonal adjustments using the U.S. Commerce Department’s standard methodology in order to enable more meaningful month-to-month comparisons. This adjustment process accounts for normal seasonal variations, holidays, and the number of weekdays in each month.
In addition to seasonal adjustments, the variables are weighted to more appropriately reflect their true contributions to freight or passenger activity. For the four freight variables, observed cargo volumes are weighted by both value and tonnage, combined equally. The six passenger variables are weighted to reflect total passenger volumes. Most of the passenger variables measure this directly, but automobile and bus traffic is weighted to account for average vehicle occupancy levels according to periodic survey data.
After seasonal adjustments and weighting, the freight and passenger variables must be combined to obtain the freight and passenger components of the PA Pulse. The six passenger variables are added together. The four freight variables are combined using a chain weighting methodology to remove the effects of inflation and better represent the real values of those freight variables in a meaningful composite indicator. Both the freight and the passenger components are indexed with a base year of 2010.
What are some factors that influence the PA Pulse?
The PA Pulse is intended to reflect the level of activity on the regional transportation system, adjusted for seasonality, holidays, and the number of weekdays and weekend days in each month. Aside from these seasonal adjustments, the PA Pulse is not further smoothed or averaged, in order to retain its ability to reflect some of the real dynamics of activity on the system each month.
A wide variety of factors influence the index. First, the index is sensitive to overall economic conditions, such as employment and consumer expenditure levels, which influence the amount of commuting and goods movement in the region. The Freight Pulse is also sensitive to phases in the business cycle that require less transportation activity (e.g., periods when businesses favor drawing down on inventory instead of restocking goods from afar).
Second, the index is influenced by the weather. Numerous major weather events have clearly left their mark on the index:
Third, the index is influenced by changes in transportation intensity over time, such as changes in bus or automobile occupancy, or average amount of goods carried on each truck. The design of the index is intended to account for these factors, but new data to update these factors is available only every few years. As a result, if any of the factors begin to trend in a new direction (such as a sudden rise in carpooling due to high gas prices), the index will not reflect these changing realities until new survey data becomes available.
Finally, the index is influenced by the ongoing dynamic activity patterns on the region’s transportation systems. The amount of interstate commuting is influenced not only by population and employment levels, but also by individuals’ ability to find work near where they live. Similarly, the amount of interstate trucking is influenced by businesses’ warehouse location strategies. Activity at the region’s ports is affected by exchange rates and the economics of international shipping.
How does the PA Pulse vary with the business cycle?
The transportation system is a fundamental component of any regional economy. It connects people with opportunities for work, education, shopping, and recreation; it conveys goods to markets; and it enables trade and travel to and from other regions around the world. In few places is the transportation system more integral to the economy than in the New York/New Jersey metropolitan region, where that region’s vast size, population, and density create extraordinary mobility challenges, and where its prominent role in international trade and travel have long been key factors in the region's and the nation's economic competitiveness.
Transportation activity offers a window into the strength of the economy. At a national level, there has been growing interest and research into the relationship between transportation activity and economic cycles. In 2003, researchers at the State University of New York at Albany showed that variations in the output of transportation services can serve as a leading indicator of growth cycles in the U.S. economy, preceding turning points in those growth cycles by an average of five to six months1. The following year, the U.S. Department of Transportation created the Transportation Services Index (TSI), which measures the output of transportation services in the U.S. economy on a monthly basis2. In this region, researchers at Rutgers University demonstrated that truck activity on the New Jersey Turnpike is a leading indicator of regional employment3. Researchers at the College of Staten Island have also explored the challenges inherent in developing a comprehensive index like the TSI at a regional level4.
While the PA Pulse resembles the TSI in a number of ways, it does not measure economic activity as directly as the TSI does. Some of the activity captured by the PA Pulse responds to economic change beyond the immediate region. Thus, the PA Pulse is better understood as a transportation activity index than as an economic index.
Although the PA Pulse clearly does rise and fall with the regional economy, it is relevant to inquire whether it has any utility as an indicator of the economy. We have started to explore this question by comparing the PA Pulse with two regional economic indicators produced by the Federal Reserve Bank of New York: The New York City Coincident Economic Index and the New Jersey Coincident Economic Index5.
To obtain a clearer comparison of the indicators, we have used a smoother version of the PA Pulse for this analysis. X12-ARIMA, the software used to seasonally adjust the components of the PA Pulse, statistically decomposes raw time series data into three basic components: a Trend component (the medium- to long-term movements of the series, including key turning points), a Seasonal component (recurring and predictable within-year fluctuations), and an Irregular component (the remaining noise due to weather and other nonrecurring factors). The PA Pulse filters out the Seasonal component to leave only the Trend and Irregular components, which combine to portray both the big picture and more fine-grained activity changes on the system. We are also able to use the Trend component by itself, which is particularly useful for a cleaner comparison between the PA Pulse and other data series.
A chart of the results appears below.
These results suggest anecdotally that either the freight or passenger component of the PA Pulse sometimes leads the coincident economic indicators, but not in a consistent or definitive way. In some cases, the PA Pulse does appear to hit a peak or trough in advance of the coincident indicators, but takes several additional months to make a definitive change in direction. In the case of the PA Freight Pulse, there are also sub-cycles of activity within each business cycle, in which freight activity trends upward or downward (perhaps due to changes in businesses’ decisions about when to replenish inventory) without similar movements being observed in the regional economic indicators.
At this point, there is not yet enough of a historical record or statistical analysis to demonstrate that the PA Pulse predicts changes in the regional economy. In the meantime, it is safe to say that the PA Pulse roughly tracks the coincident indicators of the region’s economy. We will be monitoring these indices and conducting more rigorous statistical analysis in the months ahead.
How does the PA Pulse compare with the U.S. Transportation Services Index?
In 2004, the U.S. Department of Transportation created the Transportation Services Index (TSI), an experimental economic indicator that measures the output of transportation services in the U.S. economy on a monthly basis. The TSI played a significant role in informing the development of the PA Pulse. The PA Pulse has adopted many aspects of the TSI’s methodology, and applied these to one specific region and set of facilities. There are some general similarities between the two indices: Both have separate freight and passenger components that capture many modes of activity, and both are computed and seasonally adjusted using similar methods.
Despite these general similarities, however, the two indices measure significantly different things. First, the TSI measures the economic output of the U.S. transportation services sector and relies heavily on private sector reporting, which is not practical to reproduce at a regional level. The PA Pulse focuses on the directly observable movement of goods and passengers on the regional transportation network using data obtained from PA operations, or from our direct public- and private-sector partners. As a result, the PA Pulse includes components that can be measured on the system but are not easily quantified economically, such as private automobile travel.
A second difference is that the PA Pulse focuses on a specific network of facilities, which includes a small but broad-based slice of the region’s overall transportation activity. In contrast, the TSI has somewhat narrower modal coverage but is intended to measure total activity across the entire network at the national scale—in particular, the value of transportation services contributed to GDP.
Each index includes a qualitatively different cross section of transportation modes and activity on those modes. A comparison of the components of the two indices appears below.
Components of the U.S. Transportation Services Index and PA Pulse
|U.S. Transportation Services Index||PA Pulse|
|Weighting Basis||Value of transportation services||Value and tons of goods;
Number of passengers
|Air Cargo: Domestic||Yes||Yes (4 commercial airports)|
|Air Cargo: International||No||Yes (4 commercial airports)|
|Trucks: Private||No||Yes (6 interstate crossings)|
|Trucks: For-Hire||Yes||Yes (6 interstate crossings)|
|Trucks: Parcel/Mail Services||Yes||Yes (6 interstate crossings)|
|Rail: Carloads and Intermodal||Yes||No|
|Water: Domestic Waterways||Yes||No|
|Water: International (containers)||No||Yes (Port of NY & NJ)|
|Water: International (bulk, other)||No||No|
|Air||Yes (domestic only)||Yes (4 commercial airports)|
|Automobiles||No||Yes (6 interstate crossings)|
|Buses: Transit/Commuter||Yes||Yes (6 interstate crossings)|
|Buses: Intercity||No||Yes (6 interstate crossings)|
|Buses: Tour/Charter||No||Yes (6 interstate crossings)|
|Ferry Passengers||Yes||Yes (Interstate routes)|
|Rail: Rapid Transit||Yes||Yes (PATH system only)|
|Rail: Commuter||Yes||Yes (Trans-Hudson only)|
1. K. Lahiri, H. Steckler, W. Yao, and P. Young. “Monthly Output Index for the U.S. Transportation Sector.” Journal of Transportation and Statistics, v. 6, n. 2/3 (2003). Available at: http://www.bts.gov/publications/journal_of_transportation_and_statistics/volume_06_number_23/.
2. See Bureau of Transportation Statistics, “Transportation Services Index” website, available at: http://www.bts.gov/xml/tsi/src/index.xml.
3. K. Ozbay, N. Mantel, P. Woods, and M. Robbins, “Freight Movement as an Economic Indicator for New Jersey / New York / Connecticut Tri-State Area.” Region 2 University Transportation Research Center (2008). Available at: http://www.utrc2.org/research/projects.php?viewid=102.
4. G. Wang and J. Peters, “The Usefulness of U.S. Transportation Service Index for New York State/Metro Area.” Region 2 University Transportation Research Center (2009). Available at: http://www.utrc2.org/research/projects.php?viewid=167.
5. See Federal Reserve Bank of New York, “Regional Indices of Coincident Economic Indicators,” website, available at: http://www.newyorkfed.org/research/regional_economy/index.html.
The Port Authority Pulse is a broad-based and seasonally adjusted indicator designed to provide a timely and comprehensive picture of activity on the regional transportation system. It has two components: a PA Freight Pulse based on data from Port Authority facilities and operations, including air cargo, port, and interstate truck activity, and a PA Passenger Pulse based on travel and ridership data for the region’s airports and interstate passenger transportation facilities. These indices track transportation activity on the key markets and networks served by the Port Authority -- the region’s port and airport gateways and its interstate transportation network – on a monthly basis over the past 22 years.
The PA Pulse is part of the Port Authority’s effort to make transportation activity data more accessible and useful. Three years ago, the Port Authority began reporting transportation data in its Monthly Economic Indicators report as part of its ongoing economic monitoring and forecasting program. But transportation activity indicators such as traffic volumes do not easily lend themselves to trend analysis, since they are strongly influenced by seasonal changes and calendar effects, and since activity tends to move among the various modes and facilities due to changes in user costs and service levels. An index that incorporates a broad cross-section of transportation modes and facilities, and which is statistically adjusted to account for seasonal and calendar effects, can provide a more useful measure of transportation activity that can be tracked and analyzed over time.
This document describes the methodology used to develop these indices.
The PA Freight Pulse measures goods movement activity on the interstate transportation network and key intermodal gateways, weighted according to the estimated value and tonnage of cargo transported on each facility. It combines four variables to measure goods movement at the Port Authority’s airports, port facilities, and bridges and tunnels:
The air cargo and truck variables are available on a monthly basis back to at least 1992. The port commerce variables are available on an annual basis only for 1995-2004, so monthly estimates for these years were derived from the annual totals using the average annual share of each month from the later years. The table below describes the variables used for the index in more detail.
|Air Cargo Revenue Freight||Port - Containerized Cargo Imports||Port - Containerized Cargo Exports||Truck Cargo|
|Units||Monthly short tons of air cargo (dom. + int’l) at four PANYNJ commercial airports||Monthly loaded imports of containerized cargo, in twenty-foot equivalent units (TEUs)||Monthly loaded exports of containerized cargo, in twenty-foot equivalent units (TEUs)||Monthly eastbound trucks at PANYNJ interstate crossings|
|Data Availability||1990-present||2005-present; monthly data estimated from annual totals for 1992-2004||2005-present; monthly data estimated from annual totals for 1992-2004||1990-present|
|Weighting Factors||Value per kg for international air cargo, estimated from the World Trade Atlas||Ratios of value and tons for maritime trade imports (from the World Trade Atlas and the Foreign Trade Division of the U.S. Census Bureau) to TEUs (from Port Authority data)||Ratio of trans-Hudson freight (from the Commodity Flow Survey and Freight Analysis Framework) to observed truck volumes|
|Contribution to Index as of December 2013||20.3%||24.3%||8.5%||49.6%|
These four variables were seasonally adjusted using X-13-ARIMA-SEATS (a software program developed and used by the US Commerce Department) to account for seasonal factors, calendar and trading day effects (e.g. number of weekdays in each month), and floating holidays1.
The index does not include a number of important components of the overall freight system, including domestic waterborne commerce, pipeline transportation, and rail freight transportation. It also excludes trucks that cross the Hudson River north of the George Washington Bridge, as well as other truck movements that remain on one side or the other side of the waterways separating New York and New Jersey.
Raw and adjusted trendlines for each of the four components follows.
Note: Monthly data for 1992-2004 estimated from annual totals
Note: Monthly data for 1992-2004 estimated from annual totals
In order to combine the freight variables into a single index, they had to be weighted and converted into consistent units of measure. Unfortunately, any specific choice of a measurement unit (volume, mass, value) is arbitrary, yet the choice has a significant impact on the relative weights of the different transportation modes. Using the criterion that the weighting methodology should provide a balanced perspective on the freight activity on the regional transportation system, we adopted a combination of mass and value as our weighting basis. The freight index was computed twice, once weighted by mass and once weighted by value, and the results were combined into the final freight index.
Estimates of the value per ton of international air freight were derived for the New York region from World Trade Atlas data, and seasonally adjusted. Data on the historic value of domestic air cargo at Port Authority facilities was not readily available. Based on national trends (derived from USDOT’s Freight Analysis Framework), we estimated that the value per ton of domestic air cargo has varied from about 78% of the value of international air cargo in 1997 to about 86% in 2012.
Port Imports and Exports
Port Imports were weighted according to the ratio of the value and tonnage of waterborne imports through the Port of New York customs district (according to the World Trade Atlas and Census Bureau data) to the observed number of loaded TEUs entering the port in a given month, and seasonally adjusted. Note that these are dissimilar variables, since the World Trade Atlas data includes the value and tonnage of containerized cargo as well as bulk cargo, general cargo, motor vehicles, and other non-containerized goods. Thus, the weighting factors accomplish two things: they scale containerized trade up to reflect the total trade (containerized + non-containerized) at the port, and they account for the changing value of this physical trade over time. The factors were estimated based on annual data for 1992-2004 and monthly data for 2005-2013.
Port Exports were handled in the same manner as Port Imports. Exports were weighted according to the ratio of the value and tonnage of waterborne exports through the Port of New York customs district (according to the World Trade Atlas and Census Bureau data) to the observed number of loaded TEUs entering the port in a given month. Again, these are dissimilar variables, since the weights include bulk, general cargo, motor vehicles, and other non-containerized goods. Thus, this weighting factor scales containerized trade up to reflect total trade, and accounts for changes in this total trade over time. These factors were estimated based on annual data for 1992-2004 and monthly data for 2005-2013.
Finally, continuous time-series data that would enable an estimate of the value of cargo carried by trucks using the Port Authority’s interstate crossings are not available. We developed proxies for this using two federal data sources: the Census Bureau’s Commodity Flow Survey and the USDOT’s Freight Analysis Framework.
For each of these datasets, the total value and tonnage of eastbound and westbound goods movement at the Hudson River and NJ/NY interstate screenline was estimated. For the purposes of this estimate, the “East of Hudson” region was defined as the six New England States plus the New York State portion of the New York metropolitan region. “West of Hudson” was defined as the rest of the country2
For each of the data series available, the total value and tonnage of eastbound and westbound trade was computed. Two key adjustments were made to these data:
After these factors were applied, the resulting estimates of value and tonnage using Port Authority facilities was divided by the total number of eastbound trucks at Port Authority crossings to arrive at aggregate estimates of two-way interstate goods movement per eastbound truck (since the index is based on eastbound truck movements). Trendlines were estimated from these results and were used to weight the truck data.
The PAAI Freight Index measures freight activity on the Port Authority’s system, with the components weighted by the value and tonnage of goods handled. It is important to note that this is not the same as an index based directly on the value or tonnage of goods. A value-based index, for example, would tend to rise much more sharply over time because of the effects of inflation and the improved economic productivity of freight movement due to the shrinking size of many consumer goods. We believe that an activity-based indicator is more meaningful because (1) it relies more heavily on Port Authority’s directly-measured data, (2) it is not unduly influenced by overall increases in the value of goods being shipped, and (3) it more accurately characterizes the amount of transportation activity taking place.
Following the practice used by the TSI and many indices published by the Federal Government, four variables were combined into a chain-weighted composite freight indicator using the Fisher Ideal Index methodology:
Where, in this case,
qj,t is the quantity of transported by mode j at time t; refers to the index year
pj,t is the unit price or value of the goods transported by mode j at time t;
qj,0 and pj,0 refer to the index year
Once the weighted composite indicator is developed, it is indexed to a 2010 base year (the average value in that year is set to equal 100).
The PA Passenger Pulse measures activity on the region’s commercial airports and its Interstate Transportation Network, which serves travel between New York and New Jersey. Specific facilities and transportation services included in the PA Passenger Pulse include:
This activity is weighted according to the estimated number of passengers using each transportation mode.
The PA Passenger Pulse is comprised of six variables: air passengers, interstate auto passengers, interstate bus passengers, interstate ferry passengers, PATH passengers, and interstate commuter rail passengers. Data on air passengers, PATH passengers, and auto and bus volumes are available on a monthly basis back to at least 1992. Data were not available on a monthly basis, and were estimated from annual totals, for commuter rail (1992-1995) and ferry ridership (1992-1995, 1997). These variables are seasonally adjusted, weighted to reflect the relative number of passengers they accommodate, and normalized to a 2010 base year. A summary of the data sources used for the Passenger Pulse appears below.
|Description of Variable||Air Passengers||Auto Passengers||Bus Passengers||Ferry Passengers||PATH Passengers||Commuter Rail Passengers|
|Units||Air passengers (domestic and international)||Interstate eastbound two-axle passenger vehicles||Interstate eastbound buses||Interstate ferry ridership||PATH ridership||Trans-Hudson Commuter Rail Ridership|
|Data Availability||1992-present||1992-present||1992-present||1998-present; monthly estimates for 1992-97||1992-present||1996-present; monthly estimates for 1992-95|
|Weighting Factors||None||Passengers per vehicle from PANYNJ Vehicle Occupancy Surveys||Passengers per bus from PANYNJ and NJ Transit surveys||None||None||None|
|Relative contribution to index in Sept. 2012||16.9%||40.7%||23.7%||1.1%||10.6%||7.0%|
The index represents a broad cross-section of passenger travel in the region, but it is not complete. It currently does not include General Aviation or Amtrak passengers, or passengers who cross the Hudson River north of the George Washington Bridge. It also excludes travel that does not cross one of the waterways separating New York and New Jersey.
The index is weighted to represent total passenger movements on the interstate crossings and at the region’s airports. In some cases, this requires developing estimates of the number of passengers in each vehicle. Since no time series data on automobile occupants are available, vehicle counts were expanded into estimates of vehicle occupants using data on average passengers per vehicle drawn from the Port Authority’s vehicle occupancy surveys, which were conducted in 1993, 1996, 1997, 2000, 2002, 2005, 2008 and 2013. An overall trendline for vehicle occupancy was estimated, and this trendline was used to convert the automobile count data into estimates of auto passengers. Combining vehicle counts with the passengers per vehicle estimates, and accounting for travel in the westbound direction, produces the following estimate for the monthly number of passengers traveling by auto through Port Authority facilities:
Similarly, there is no comprehensive, time series data are available on the average number of passengers per bus. We have estimated this from two sources. First, we have used data from the 1993, 1994, 1995, 1996, 1997, 1998, 2004, 2008 and 2011 Continuous Bus Surveys, which measured the occupancies of buses arriving or departing from the Port Authority Bus Terminal and the George Washington Bridge Bus Station, to establish average bus occupancies at these facilities. Applying these occupancies to the entire population of buses crossing Port Authority bridges and tunnels may introduce errors for a number of reasons:
Despite these problems, the Continuous Bus Survey is the most comprehensive data source available for estimating these passenger volumes. Since some of the errors mentioned above distort the estimated bus occupancies in opposing directions, it is possible that they partially cancel each other out.
The variations in bus occupancies over time were based on monthly bus data provided by New Jersey Transit for its trans-Hudson bus operations. NJ Transit represents a significant share of all trans-Hudson bus operations, but a key assumption here is that changes in their passenger loading levels are generally reflective of the market as a whole.
Applying these changes to the Port Authority’s broader bus occupancy estimates, and its historic bus counts, accounting for westbound ridership, correcting for spurious fluctuations in minibus counts, and applying seasonal adjustments, produces the following estimates of total bus passengers:
The number of passengers using the trans-Hudson ferries, the PATH system, and the region’s commercial airports are all directly available on a monthly basis. Beyond seasonal adjustments, no additional weighting was applied to these data. Graphs of the trend lines for these data sources appear below.
Note: Monthly data for 1992-1995 estimated from annual totals
Note: Monthly data for 1992-1995, 1997 estimated from annual totals
All components of the PA Passenger Pulse are already represented in equivalent units, so no weighting to a comparable basis is necessary. The Passenger Index is simply the sum of the six components discussed above. No chain weighting is required because the index is a direct measure of activity and does not need to take prices into account. The index measures change from the 2010 base year.
In April 2013, the first annual revisions were made to the PA Pulse. Key changes included:
In April 2014, the second annual revisions were made to the PA Pulse. Key changes included:
The Port Authority Pulse combined indicator is computed as a simple average of the PA Freight Pulse and PA Passenger Pulse. Taken together, these indices are a first step toward enhancing understanding of major trends on the regional transportation system. They are intended to provide an aggregate barometer of activity trends on the Trans-Hudson and gateway markets in terms of passengers and freight flowing through the system. They are not a direct measure of activity or sufficiently fine-grained to serve as a basis for performance measurement or management decisions.
The Port Authority is at the beginning stages of exploring this index, how it can be used, and how it relates to regional economic cycles. Our intent is to publish this index on a monthly basis, recalibrate it annually in the spring, and engage in an ongoing effort to monitor its relationship to the evolving regional economy.
1. See U.S. Census Bureau, “The X-13-ARIMA-SEATS Seasonal Adjustment Program” website, available at: http://www.census.gov/srd/www/x13as/.
2. This introduces a few minor errors, such as the inclusion of Rockland County as “East of the Hudson” and Rensselaer County as “West of the Hudson,” but these errors are very small compared with the totals.