Canada’s airports: Their role as cargo gateways

Archived Content

Information identified as archived is provided for reference, research or recordkeeping purposes. It is not subject to the Government of Canada Web Standards and has not been altered or updated since it was archived. Please "contact us" to request a format other than those available.

Canada’s airports: Their role as cargo gateways

Using just about any yardstick (tonnage, value of goods, revenue earned, tons/kilometres), it is difficult to dispute the recent growth in air cargo traffic. Factors ranging from industrial globalization to business practices like "just-in-time" inventory management and aggressive off-shore marketing of Canadian goods and services have contributed to this increased activity. As local groups contemplate the purchase or lease of community airports, they are being advised to secure a solid cargo base in addition to passenger traffic, the rationale being a more diverse revenue stream for the airport as well as a leg-up for local industry resulting from the economic stimulus derived from air connections.

In the past seven years, air cargo traffic enplaned and deplaned at Canadian airports has increased at about 3 percent per year, on average. (See Figure 1.) Breaking this total into its components, it appears that international freight has shown the most improvement in terms of tonnage. (See Figure 2.) The cessation of Air Canada’s dedicated DC-8 freighter service in 1994 hit the domestic side of the picture, although other carriers have recently increased cargo service in Canada.

Looking at international airfreight from a different point of view, it can be demonstrated that imports and exports transported by air have also grown, at a rate generally exceeding the overall growth in international trade for Canada. (See Figure 3.) In this instance, total airfreight is expressed in terms of value, whereas the previous two illustrations presented traffic in terms of tonnage. A detailed examination of the trade database is presented later. For the time being, it can be seen from the gains made in the total value of merchandise shipped by air that more products appears to be moving in this manner.

A third indicator of goods transportation is the financial and operating information filed by Canadian air carriers (the following illustration includes only Canadian carriers, whereas the two previous ones included foreign carriers serving Canadian airports). Cargo service represents about 10 percent of total operating revenue, however, carriers see the value of attracting cargo business in terms of better utilization of capacity. In fact, about 85 percent of goods transported by air are carried as bellyhold cargo on scheduled or charter passenger flights; dedicated cargo flights account for the remainder. As a recent promotional pitch aptly phrased it, "... you shouldn’t fly on an empty tummy".

Canadian carriers saw their overall revenue picture darken during the recent recession. However, despite the disappearance in 1994 of the Air Canada freighter service, cargo revenue had almost regained its pre-recession peak by the end of 1994. In fact, the first half of 1995 saw goods transportation revenues up 8 percent over the same period in 1994 (total operating revenue increased by 6 percent).

Figure 4 shows goods tons/kilometres performed by Canadian Level I-IV carriers since 1988. Overall ton/kilometres grew by nearly 10 percent in 1993 and 1994, following two poor years. Decomposing the total reveals a sluggish progression for domestic cargo and more impressive growth in the trans-border and other international sectors. Preliminary results for the first half of 1995 show overall goods tons/kilometres up by 13 percent over the previous year. Domestic cargo grew by 7 percent, while other international freight was up 10 percent and trans-border freight surged ahead by 66 percent.

The preceding illustrations depict what has been going on in Canada. To get a picture of what is happening the world over, it is useful to examine statistics produced by the International Civil Aviation Organization (ICAO). ICAO reports show overall improvement for carriers at the world level. (See Figure 5.)

In a repeat of the pattern exhibited in Canada, international cargo output has grown at a faster rate than domestic cargo tons/kilometres performed in most countries. The size of many ICAO countries would make surface or perhaps even water transportation competitive with air transport for domestic freight. This may be a contributing factor to the superior growth of international cargo traffic.

Even though each of the preceding illustrations tell essentially the same story, the application of any of these sources on their own would be of limited use to someone who wanted to study cargo markets in terms of what is being moved and where. In order for a local economic development team to determine whether a particular airport operator would be well advised to pursue certain types of cargo, or for a shipper to evaluate the feasibility of locating a plant, storage facility or sales and support centre in a particular region, reliable, timely information is required on what is being produced and shipped to and from centres of interest.

The two principal data sources that lend themselves to this type of study are the Airport Activity Survey, produced by the Aviation Statistics Centre and the International Trade data series, produced by the International Trade Division, both at Statistics Canada. Neither currently provides a complete picture of what is going on at Canadian airports. However, the next section details how a useful composite data source for international air cargo might be developed using these two databases.

This is actually a collection of surveys covering a range of Canadian and foreign carriers using Canadian airports. The principal target population comprises major scheduled carriers and major charter carriers. Passenger information is also collected from regional and local carriers, but they do not file cargo data.

The major scheduled carriers file a report detailing the daily arrival and departure of every flight, including seats available, cargo capacity available, and passengers and cargo on board. Monthly reports are produced from this information and released on a quarterly basis. Major charter carriers also report on a monthly basis; they file less detail, but they do report cargo information. These are census surveys, i.e. all carriers in these two categories are required to file this information.

Passenger and cargo traffic are calculated on an enplaned/deplaned basis for each airport (space restrictions permit the publication of only the top 50 or sometimes top 100 airports, but data can be obtained for airports not listed in the published reports). Both entity cargo (dedicated cargo flights) and bellyhold cargo (the use of space in the cargo hold on passenger flights) are filed for scheduled and charter flights.

The "to and from" part comes in with the reporting of route-related information, which is derived from the reported routing of the aircraft (next stop for departures, last stop for arrivals), although the derivation of the ultimate origin and destination of the cargo is not possible. For example, a flight with a routing of Vancouver-Edmonton-Amsterdam: "... the departure of the flight from Vancouver would be considered domestic, as would the arrival at Edmonton. The departure of the flight from Edmonton for Amsterdam would be considered as other international since the next stop of the aircraft is a foreign point."

Using the next/last stop information, the Airport Activity Survey calculates, by airport, the tons of cargo loaded and unloaded for domestic, trans-border and other international flights. Since the journey of the cargo may not coincide with the flight segment (and may indeed require more than one flight), the "true" origin and destination of these goods cannot currently be inferred from this data source. An examination of the survey output for major scheduled carriers shows the presence of a "flight destination" field, which often differs from the next/last stop field, but which does not figure in the present published output. This is the final origin/destination of the flight, which could give a better approximation of the "true" origin/destination of the goods on board.

The flight origin and destination will give more information on cargo for some flights, specifically those with one station stop (as in the Vancouver-Edmonton-Amsterdam example). The end points become less relevant on flights with several stops, where cargo is enplaned and deplaned at each stop. While there may be many non-stop or one-stop international flights, this aspect of the data needs further study in order to determine the quality of what is reported, and potential application for producing O/D information. The remainder of the traffic (the multi-stop "milk runs") would require alternate treatment, possibly the development of a model.

There is no detail reported on the goods carried, nor is the value of the cargo reported on the Airport Activity Survey. The financial data reported by Canadian carriers who remit financial statements to the Aviation Statistics Centre relate to the revenue earned from the transportation of goods; no information on the value of the cargo itself is captured at the Aviation Statistics Centre.

In terms of the coverage of the Airport Activity Survey, only Level I, some Level II Canadian carriers and equivalent foreign carriers file cargo data. The absence of information on regional and small carriers results in an understatement of total cargo activity. Civil Aviation operating data indicate that about half the tonnage enplaned by Canadian carriers is not captured in the Airport Activity Survey. Only passenger traffic detail is required of the smaller scheduled carriers; the smaller charter carriers don’t file traffic data at all.

Other aviation surveys can help in approximating the understatement of cargo activity in the Airport Activity Survey. If the objective is to derive cargo flows and not to account for the total (i.e., unduplicated) volume, it would not be inappropriate to show goods moving twice if that is in fact what happens. For example, goods coming in from the Pacific Rim to Vancouver may be delivered by a national carrier to an airport further inland, then fed to a regional or local carrier for shipment to their destination. Ignoring the last leg of the journey for these goods will not result in an accurate O/D, thus distorting cargo patterns. The same can be said for intermodal transfers (marine/air, air/truck, etc.) which have not been considered in this paper, but which would be necessary to really know the ultimate origin or destination of these goods.

Following on the earlier statement that most cargo is not carried on dedicated cargo flights (insofar as major scheduled carriers are concerned), if the same holds true for smaller carriers (who presumably would also want to maximize capacity, given the opportunity). Using Aircraft Movements Survey (AMS), counts of arrivals and departures for Level I to III and smaller carriers at selected airports show gaps between AMS and the Airport Activity Survey. (See Table below.)

Granted, the aircraft movements counts do not distinguish full from empty aircraft and passenger from cargo flights, but the apparent undercounting can be seen by the relative numbers of flights observed in both surveys.

To sum up, the Airport Activity Survey has some elements of origin and destination information, some of which have yet to be investigated. The presence of a flight O/D, in addition to the next/last stop data, could be used to produce a better approximation of some cargo movements (non-stop or one-stop flights, as noted earlier). The positive aspects of this database include full coverage of movements by major carriers (Canadian and foreign). The negative points to consider include the absence of information on the description and value of the goods, and the gap resulting from the fact that regional and local carriers are not obligated to file goods-related information.

Statistics Canada compiles an extensive set of information relating to imports and exports of goods. This major economic series is derived from administrative data from Canadian and American customs and excise records. Under a Memorandum of Understanding (MOU) between Canada and the United States, the two countries "swap" their respective import data for the other country. Thus, Canada receives a database detailing American imports from Canada, which are then expressed as Canadian exports to the United States. This exchange (the MOU was signed in 1987, which began in 1990, has resulted in access to additional data elements from the US file which are not available from the standard export document (B13). Exports to other countries are derived from B13 reports, and all imports are derived from the B3, both of which are Revenue Canada documents.

Trade data are compiled monthly by country of origin (destination), and by type of commodity (using the Harmonized System of classifying goods). Other information available include the value of the merchandise, the province of origin/destination and the mode of transport. The mode of transport (MOT) refers to the means by which the goods left the country for exports, and the mode by which goods arrive in Canada. Most international airports are also Customs ports of entry/exit, and a major assumption in the following analysis is that most merchandise transported by air clears Customs in either a Canadian or a foreign airport, as opposed to a sufferance warehouse (in which case air freight would then be categorized as road).

The Trade file contains the date (day, month and year) of the shipment, the eight or 10-digit commodity code, the Canadian port of entry/exit and the country of origin (and the country of export for imports. Other fields include the quantity and a unit of measure (since some commodities are shipped in units other than kilograms); quantities are converted to kilograms, where appropriate. The value of the goods is also given, as well as the identity of the importer or exporter, and the identity of the carrier (sometimes). For US commerce, the origin and destination are given at the state level, and the shipping weight is required at the "line level" for marine and air shipments valued over $2,500 (as opposed to "document level", which is the requirement for other exports).

All merchandise entering or exiting Canada must clear Customs. The recent appearance of Foreign Trade Zones (FTZ) near some airports would be an exception. Goods may enter a FTZ in bond, where a manufacturer would re-assemble, label, re-package or somehow modify the product, then ship it out of the country. There are few FTZ’s in Canada - Vancouver and Edmonton have them, Hamilton is developing one. They can be expected to proliferate as industry continues to segment production and distribution of materials, according to where expertise and efficient labour and facilities can be found. The growth of this activity will have an impact on the use of Trade data to approximate airport-related activity, since these goods will flow through the airports (twice) but not through Customs.

Analysis of goods movement using Customs data for shipping and trucking has been problematic when goods movements are being reconciled between transportation and trade surveys, since product can come into the country and go into bonded storage for up to months at a time. A second assumption in the aviation analysis is that this factor is minimal, because of the general time sensitivity of airfreight. This is not to say that no air freight waits around in a warehouse. It is likely, however, that the storage period would not exceed a month (and even less likely a quarter) for time sensitive material.

The following evaluation of the trade statistics for airfreight, in comparison to the Airport Activity international cargo data, will demonstrate whether it is appropriate to combine these two data sources to make statements about the flow of goods by air. Coverage and composition of the data and the underlying concepts of each survey are factors, which will be considered.

The initial study period was one quarter, but this was narrowed down to July 1994. In order to evaluate import and export data for the air mode, all records coded to air for July 1994 were extracted, which resulted in four databases, two each for imports and exports (US and other). Detailed output from the Airport Activity Survey was also reviewed for July 1994.

Six Canadian airports were selected, based on their importance to the overall international cargo scene: Dorval; Mirabel; Toronto; Hamilton; Calgary and Vancouver. These airports appear regularly (with the exception of Hamilton) in the quarterly Airport Activity tables, and all are served by a combination of major and regional or local carriers. A summary of the results for both surveys for July 1994 is presented below.

At a glance, Toronto dominates in terms of tonnage as well as value. Hamilton barely registers on the Airport Activity Survey (most of the tonnage reported is classified as domestic, possibly due to routing of the flights), yet there is about $70 million in goods moving through this port of clearance, just for one month. No American-bound cargo was recorded out of Hamilton, yet nearly two-thirds of inbound cargo (in terms of value) at Hamilton arrived on flights from the US. Vancouver placed second in the Airport Activity results on a tonnage basis, but Mirabel outdistanced Vancouver on a value basis for two-way trade.

Toronto - major destinations were in the US (California, Texas account for 13 percent each of value of shipments; principal cargoes include precious metals, computers and other electronic equipment); 20 percent of inbound cargo is shipped from California (although many shipments of electronic equipment originated in Asia). Important cargoes unloaded included machinery and pharmaceuticals, precious metals, electronic goods and motor vehicle parts. Relatively low-value goods included plastic and rubber goods and textiles.

Mirabel - one-quarter of shipments were destined for Great Britain, another 17 percent for Switzerland; Britain and Germany were important points of origin for international air shipments at 13 percent each. Important commodities moving through Mirabel included aircraft and parts, chemicals and pharmaceuticals and precious metals.

Dorval - given the composition of airlines calling at Dorval in July 1994, it was not surprising to see American airports come out on top in terms of air cargo shipments in both directions. Mexico was also an important destination. Important high-value cargoes included aircraft and parts, computer-related equipment, pharmaceuticals and chemical products.

Vancouver - nearly one-third of inbound cargo was shipped from California - many shipments originated in Asia. Computer equipment, optical fibres and aircraft were important commodities delivered to Vancouver. Japan and Hong Kong were important offshore destinations; they accounted for 27 percent of shipments by value. Precious metals and computer equipment were important commodities; also registering significant dollar values were shellfish and ginseng bound for Hong Kong.

Calgary - nearly one-quarter of shipments into Calgary came from California, many of these were electronic goods which originated in Asia. Low-value cargoes included rubber and plastic products, some of which originated in Taiwan. Outbound cargo included telecommunications equipment for Britain and chemical products headed for California.

Hamilton - California and Italy each accounted for 13 percent of inbound cargo. Precious metals and electronics equipment were high-end commodities; plastic and leather products and printed material were relatively low-value commodities entering Hamilton from these two markets. Heavy construction equipment was an important outbound cargo category, significant quantities were destined for Brazil.

Trade data are based on administrative forms covering virtually 100 percent of goods shipped into and out of Canada. The Airport Activity Survey also covers 100 percent of its target population, which is nonetheless restricted to major air carriers. Some sort of link was sought which would show the relative coverage of these data sources, obvious candidates being quantity, time period, origin and destination. Links were also sought on carrier code, but this was not available on the export file.

The best link between the two data sources would have been the weight, but most trade records extracted had no entry in this field. Information on quantity is collected only for goods where the Customs Tariff and export classification deem it necessary. The value of the goods drives the calculation of duty, where applicable, and it is also a critical variable for calculating the balance of trade between Canada and other countries.

Quantity is of lesser value (keeping in mind that the source is an administrative record collected for purposes other than describing cargo flows) to trade analysts if the value is present.

For July 1994, there were nearly 167,000 records in the import file, about half of which had no value in the quantity field of the extracted file (see below). For exports (some 36,000 records), the absence of quantity data was more pronounced. As mentioned earlier, quantity data is not required for certain types of commodities (for example, aircraft are measured in pieces). While it would be useful to examine a few more months’ worth of trade records, it would seem that quantity is a variable that would be difficult to obtain on a continuous basis.

While value was consistently present, it was difficult to substantiate the value of different types of cargo, since the quantity was not always available. Where quantities were present, a series of unit value calculations by commodity type were performed. Interestingly, not all the commodities chosen for unit value calculations appeared to be high-value merchandise (although there were several shipments of "hi-tech" materials, precious metals and electronic equipment noted); perishability appears to be a common characteristic of air freight (notably, food products) as well as time-sensitivity (machine parts). While this is not news to those who analyze aviation data, a closer look at the vast range of products that are shipped by air would likely reveal several niche market possibilities for local airport authorities, or even air carriers, producers and shippers.

Although it was not always possible to identify a carrier from the Trade data source, the more detailed import file carries a Cargo Control Number (CCN), where a carrier code is embedded. Though not always available (sometimes the carrier code is a default value), it also provides a potential link to the Airport Activity file. Knowing the frequency of service of a carrier between two points (the total number of flights for a given month are provided in the Airport Activity file for each routing), it may be possible to isolate all known records from both sources for a mechanical match. A small number of manual matches were made between the Airport Activity and Trade files for particular carriers. Few of these were successful - the foreign points were often not served by scheduled flights, according to the reference material used at ASC (mainly the Official Airline Guide), and less detail is available at ASC on charter cargo. The export file obtained via the exchange with the US includes a carrier code for southbound air and marine shipments, but this is dropped when the US data is merged with the other exports. The carrier code would be a useful tool for reconciling these two datasets.

Most international airports have their own Customs facilities for clearance of cargo, although some of these do not have the capability to clear passengers. As a result, Mirabel, Pearson and Vancouver International are all recognized Customs ports, with their own geographic codes (i.e., as opposed to "Montreal", "Toronto", "Vancouver"), which enables an analyst to place inbound or outbound cargo at an airport, thus validating the assignment of the mode of transport (MOT) code. Since only six airports were examined in detail, no statement can be made as to the overall quality of the MOT indicator. Since the "other" mode is fairly high, it would be useful to examine it in detail before deciding to use only transactions coded to air. The "other" mode transactions could be scanned for the port of entry/exit; any records showing goods moving through Canadian or foreign airports should be followed up in greater detail.

There were two dates carried on the MOU file detailing exports to the United States, showing the date of exportation and importation. Also available, but not obtained in these extractions was the date of shipment versus the date of Customs clearance. A comparative study of these would be interesting, since this study assumed (but did not prove) that air freight would likely clear Customs in a shorter interval than goods shipped by other modes.

The principal topic of this paper was international cargo traffic. The International Trade data source has some promising characteristics in relation to aviation survey data for cargo movements. Unfortunately, there is no similar companion dataset on the domestic side (existing either in commerce or aviation), although periodic surveys of manufacturers may prove useful in analyzing domestic flows.

Similarly, cargo data is not required at present from regional and local carriers - this is primarily an issue of respondent burden. Any additional requirements of air carriers must be either legislated through carrier filing requirements (by Transport Canada or the National Transportation Agency) or respond to a demonstrated requirement under the Statistics Act. Given the low priority on the part of major users (as noted above for domestic cargo), these data will not likely become part of filing requirements in the near future.

Other sources of cargo data exist which could also be compared to the Trade and Airport Activity sources. The On-Flight Origin and Destination (OFOD) and Traffic by Flight Stage (TFS) publications produced by ICAO would cover major international routes (in the OFOD series published routes must be served by at least two carriers and countries). While helpful, these statistics would not be complete. They are very useful in specific types of analysis if the above restrictions are taken into consideration.

The International Trade data source is developing as an increasing number of shippers and carriers adopt Electronic Data Interchange (EDI) for filing commercial documents. Staying on top of these developments will ensure that any improvements to this data source could be incorporated into the above analysis. The proliferation of standard EDI messages is expected to make the provision of this information less onerous on the part of the shipper or carrier.

A periodic, detailed micro match of air mode trade records and Airport Activity records should be done to compare coverage, but any ongoing O/D data program should be based on the aggregate statistics and the relative coverage of the two sources duly noted. This would save resources, as the two databases will never be the same.

1 Taken from remarks made by Dr. Michael Tretheway at the "Airports Canada 1995" conference, Toronto, December 1995.

2 Aviation Statistics Centre, Statistics Canada - Airport Activity Survey: bellyhold and entity cargo accounted for 84 percent and 16 percent, respectively, of the total tonnage in 1994.

3 Taken from remarks made by Ms. Clare O’Brien, Marketing/Advertising Manager, McCarran International Airport, Las Vegas, at the "Airports Canada 1995" conference, Toronto, December 1995.

4 Taken from notes on data quality and methodology, "Air Carrier Traffic at Canadian Airports" Aviation Statistics Centre, Statistics Canada, catalogue 51-005.

5 Canadian carriers are classified into reporting levels for statistical purposes, according to passengers carried and/or revenue earned from flying services. Level I carriers transport at least 1 million revenue passengers or 200 000 tons of revenue goods annually. Level IV carriers earn at least $500,000 annually from the provision of air services. The Civil Aviation Survey obtains financial and operating data for Level I-V scheduled and charter carriers. The Airport Activity Survey is completed by scheduled Level I-IV carriers and by major charter carriers (mostly in Level II) only, and by foreign carriers of equivalent size. The major scheduled carriers file the report covering daily traffic, Canadian Level I and some Level II carriers, and foreign carriers of equivalent size. Please refer to Statistics Canada catalogues 51-206 and 51-005 for further detail.

6 Taken from remarks made by Dave Frank, Horizon Pacific Ventures Ltd., at the "Airports Canada 1995" conference, Toronto, December, 1995.

7 Shipments valued at less than $2,000 are not captured from the B13 document. Goods entering Canada in transit to other countries are accounted for, but are not included in the import and export data presented in this paper.

8 Quantity (and unit of measure) is reported at either the "line" (commodity) or "document"(shipment) level, depending on the direction of the movement. Exports (B13) require document-level detail for all modes of transport. In this instance, the weight reported may encompass several types of merchandise, and no commodity-level quantity data will be available. For imports from the US, the line-level shipping weight (in kilograms) is required for marine and air shipments valued over $2,500.