Introduction

The contribution of transport ^{note 1} to the economy can be assessed through measures of economic performance. The economic performance of a sector improves when the value of services to the users rises more rapidly than that of the inputs required to produce them. Three conditions - or a combination of the three - are required to achieve this objective:

1. Productivity improves.

2. Productivity gains exceed changes to the price of inputs so that unit costs ^{note 2} drop.

3. Productivity gains (or the reduction of unit costs) are returned to users in the form of lower prices (this occurs in competitive and efficient markets).

Methodology

The measurement of price changes, productivity, and cost per unit of output ("unit cost") is fraught with difficulties. A rigorous analytical framework must be devised to integrate the notions of price, output, input price and input utilization. The framework used here is called total factor productivity (TFP), which expresses the relationship between the index of growth of aggregate output and that of aggregate inputs.

TFP is a more robust and comprehensive indicator of performance than is the more familiar concept of labour productivity (output per employee) in that it can reflect not only improved overall efficiency of production, but also factors such as the substitution of equipment for employees or the use of outside components or services. In other words, methods that look only at output per unit of labour input provide an incomplete picture. While output per employee might appear to improve, overall costs of production might improve much less, or even rise, in areas ignored by such methods, and this would not register in the analysis.

As the term implies, "total factor" productivity takes into account changes in all the inputs. The analytical challenge is to combine changes in the different types of inputs - labour, capital equipment, materials and services, and so on - into a single aggregate measure. The calculation is complicated by the fact that the relevant published accounts do not provide physical measures for each input - only the amounts spent on them. To overcome this problem, an index was created for each input, dividing the amount spent on it by its price per unit ("input price"). The resulting indices are then combined into an aggregate index weighted by the share of each input in total costs. ^{note 3}

Similarly, the technique combines different types of services provided by an enterprise to produce a meaningful single measure of output. Conventional aggregate measures such as tonne-kilometres and passenger-kilometres do not allow for differences in service characteristics such as multiple commodities, or differences in shipment size, haul distance, stage length or markets served. The use of aggregate output indices supposes that a coal movement costing $20/tonne is equivalent to a movement of automobile products that costs $150/tonne. Or it gives the same weight to a passenger movement of less than 450 kilometres that costs between 25 and 30 cents per kilometre as to a trans-Pacific flight of 7,000 kilometres at a cost of $0.09/km. And of course the output measure for the many carriers that provide both freight and passenger (and possibly other) services must somehow take both into account.

Again the solution is to use indexing procedures. An individual index is created for each service by dividing its revenues by its price, then the indices are combined into an aggregate output index weighted by the proportions of total revenue contributed by each service. The advantage of this approach is that it allows the calculation of comprehensive output measures of freight or passenger services - or indeed of any other services produced by transport firms. This approach also enables analysts to test the consistency of output and price measures from the following identity:

Figure 15-1 compares the output index obtained from disaggregated data in rail with the revenue tonne-kilometre index. Not only is there a 10 per cent difference between the two sets of figures, but the turning points are different. This is confirmed by the weak correlation between the two series (53 per cent).

R=PxQ

where
R= revenues;
P= price;
Q= quantity

and
P can be derived from R/Q
Q can be derived from R/P

The tonne-kilometre index shows higher growth than the output index adopted in this study. This means the estimated productivity gains would be higher, should the TKm index be used. Documented evidence attests to the magnitude of errors that can be introduced by the use of aggregate data in the analysis of rail services. ^{note 4} Likewise in the air industry, the use of aggregate output measures would yield a productivity growth nine per cent higher than results when using the disaggregated output approach.

Developing aggregate input measures is a daunting task. Inputs usually consist of multiple factors (labour, capital, fuel and other intermediate goods or services) that cannot easily be added together. This explains, in part, the attraction of calculating labour productivity, for which the amount and price of input are easily accessible.

However, labour productivity, while meaningful, remains a partial measurement. Labour may represent less than 35 per cent of total transport costs. An examination of the sources of labour productivity growth shows the limitations of relying on such measurements and ignoring the efficiency of using other resources.

There are a number of ways to improve productivity. Firms can install new technologies using more modern equipment or implement management techniques that make better use of the skills of their employees. Often, economies of scale can lead to productivity gains through increased production levels. In transport, however, economies of scale either do not exist or have been exhausted (in rail, for example). Finally, firms can achieve better labour productivity by contracting work out ("outsourcing"). The best example of outsourcing is the trucking industry: subcontracting to owner-operators rose to 25 per cent from 15 per cent between 1981 and 1994.

For this analysis, the aggregate of the indices of individual inputs was calculated by combining them and weighting each input factor according to its respective share of total costs, based on the T”rnquist approximation to the Divisia Index. As with output, indices of volume and prices of input can be tested and derived from the following identity:

C= XxF

where
C= cost;
X= price of inputs;
F= volume of inputs;
and
X can be derived from C/F
F can be derived from C/X

The TFP techniques applied here require data on cost structure (spending on each type of input) and on revenues by type of service. They also require the analyst to estimate a series of prices for both their inputs and their outputs, by type of service.

Limitations of the Analyses

Analyses of prices and output, like those of productivity and unit costs, require access to disaggregated data that are not evenly accessible. Firms operating in the same industry often have heterogeneous products (such as freight and passengers), and may each use a different mix of input components (labour, fuel, material, capital, and so on). Ideally, it is desirable to have refined historical data on all input and output components by item at the company level. Moreover, the data should be expressed in uniform and consistent units of measurement for all firms and for the entire period under review. However, this is not always possible.

A major data limitation is inconsistent data for input and output factors within and between modes. Different data base definitions, assumptions, coverage, disaggregations, reference years and calculation procedures may produce significantly different results and may influence observed trends in productivity. In two industries in particular - the inter-city bus and the marine transport industries - output and price measurements were unstable. Results for these industries are presented, but caution is in order when interpreting such trends in productivity and price changes.

All figures and tables in this chapter are from internal Transport Canada analyses. More detail on data sources and the scope of analyses is provided in Table 15-1A, in the appendix to this chapter.

Source tables for the following sections are also included in the data compendium of this report.

Output and Price Changes in Transport

Domestic Passenger Services

In 1995, the domestic intercity passenger services of the larger carriers generated $4.7 billion in revenues. ^{note 5} After exhibiting an average revenue growth of 3.2 per cent over the years, the market returned in 1995 to its peak 1989 level. The airline industry's share, currently 73 per cent of the market, is rising at the expense of rail. Since the mid-1980s, however, revenues have been growing only because of price increases, since output (or demand) is in decline. Year-to-year changes in price and volume are shown in Figures 15.2 and 15.3.

The bus industry has been showing growing revenues over the years, but output has been declining, particularly in the first half of the 1980s, when prices of services shot up 6.5 per cent per year, compared to 3.6 per cent in Canada's general economy. In the second half of the decade, price increases slowed to 3.8 per cent per year. The volume of activity also showed a downward trend for most of the period, ending up 10 per cent lower in 1995 than in 1986. Overall, the industry's market share is in the 10 to 11 per cent range.

The decline in output experienced by the inter-city bus industry is the result of price increases above gen-eral inflation trends in the economy, combined with the long-term decline in bus use, as traditional users become more affluent and rely more on private vehicles and on air travel, with its steadily declining prices.

Revenues from marine and ferry passengers, including estimated subsidies, reached almost $300 million in 1994, with government-based carriers contributing, on average, 85 per cent of this activity. Annual price increases have hovered around five per cent since 1981. The market share of the bus mode is in the five to six per cent range.

Price increases in the domestic air market in the first half of the 1980s were slightly above those in the general economy. In the second half of the 1980s, when domestic air prices advanced at about the same pace as the economy, the output of domestic air services initially increased, stimulated by improving service levels. After 1990, however, demand fell, because of the recession and rising production prices. Output tumbled for three consecutive years, to a low 19 per cent below 1990 levels. Since 1993, price changes have been moderate (less than one per cent per year) and demand has picked up (more than five per cent per year). Preliminary data for 1996 suggest a robust (eight per cent) growth fuelled by lower prices.

Revenues from passengers and government subsidies to Via Rail peaked in 1989 at almost $1 billion. Then services were severely curtailed (by half) in 1990. Since 1991, Via revenues have been relatively stable at less than half a billion. Via's market share of domestic passenger revenues has dropped from 19 per cent in 1981 to less than 10 per cent in 1995.

As a result of the reduction of subsidies, the prices paid by the users of Via services went up 5.7 per cent per year between 1981 and 1995. Price increases have slowed in recent years in spite of declining subsidies, as cost recovery improved (to 33 per cent from 25). Figure 15-4 clearly shows strong unit cost increases until 1990, followed by a steady drop (24 per cent), which has helped to diffuse the effect of the reduction of subsidies.

International Passenger Services

This market is dominated by services provided by the airline industry with only a small proportion from marine carriers. Cruise services provided by foreign carriers are not included.

The market share of charter services exceeded 30 per cent in the early 1980s and fell to less than 20 per cent by 1995. This drop was due mainly to the reduction of charter services by scheduled carriers. Price and output changes for air services are depicted in Figure 15-5.

Canadian carriers' revenues from international air passenger services reached close to $3.5 billion in 1995, a growth of 6.3 per cent per year since 1981. Demand for international passenger services has been fueled by price increases remaining below the rate of inflation in Canada. These lower transport costs represent a reduction in the real cost of travel abroad since 1981. As a result, output growth has been robust, advancing at 4.6 per cent per year, except for the major hiatus in 1991. The recession and the Iraq-Kuwait crisis that year caused a 12 per cent drop in demand and a five per cent price increase. However, 1995's strong output growth (19 per cent) pushed the industry over 1990's peak. Assuming that Canadian carriers have maintained their market share, output growth for 1996 is estimated to be in the 10 per cent range.

Given the complexity of measuring price changes in the air industry, different approaches have been used to develop indicators. Statistics Canada has produced price indices for both the domestic and international markets, using "chained Laspeyre" indices. For various reasons that have to do, among other things, with consistency between modes, another set of indices has been developed in Transport Canada (see Figure 15-6). The close parallel between the two series attests to the quality of these estimates. The slower growth of Transport Canada's price increase estimates arises from the fact that Statistics Canada treats economy, discount and deep discount fares as different services, whereas Transport Canada essentially blends them into one category. Had Statistics Canada's series on prices been used, the estimated productivity of the airline industry (discussed later) would have been lowered by 13 per cent.

Freight Services

The freight activity of larger Canadian carriers generated close to $19 billion ^{note 6} in revenues in 1995. Revenue growth reached five per cent per year in the first half of the 1980s, then remained relatively stable in the $15-16 billion range between 1988 and 1993. Growth then surged at a rate of 17 per cent in the last two years. Since 1981, overall growth has averaged four per cent per year.

Between 1981 and 1986, freight transport prices increased by three per cent per year, 0.6 per cent below general inflation trends. (See Figure 15-7, which shows annual price variations by mode.) The gap between freight and economy prices then widened at an accelerated pace, at a rate 2.8 per cent below inflation in the economy. In particular, rail freight prices declined in nominal terms at an average annual rate of 0.8 per cent (3.8 per cent in real terms). The available evidence suggests that deregulation in the transportation sector has shifted the trend line of transport prices downward, with a greater impact on rail than on trucking. Rail prices were also affected by market factors other than deregulation, such as the decline of export commodity prices between 1987 and 1991.

The freight revenues of Canadian-domiciled marine carriers reached an estimated $1.3 billion in 1995, about the same as in 1989. The considerable growth in industry revenues that year was due mainly to the capture of data relating to a large carrier that was previously missing from the survey. Between 1981 and 1988, price increases were limited to less than one per cent per year, and output dropped 1.8 per cent per year. After 1989, price changes accelerated to 3.8 per cent per year, and output fell 3.4 per cent per year (see Figure 15-7). Over the period 1981 to 1995, price changes in marine transport edged those in the general economy by a small margin, but output contracted by some nine per cent (19 per cent since 1989). This decline reflects the inertia in the major markets served by the industry, particularly in the Great Lakes (domestic and transborder traffic).

In 1994, CN and CP handled record freight volume: a 10 per cent increase in traffic over the previous year and a 6.5 per cent increase over the previous record. Traffic fell three per cent in 1995 and is estimated to have remained at that level in 1996.

Revenue growth in the early 1980s came mainly from price increases. Then railway output volume ballooned from 1986 to 1988, producing record revenues. However, this surge was followed by a significant decline in both outputs and prices, affected by both the domestic recession and the depressed prices of major export commodities.

Not until 1994 were the railways able to achieve significant growth in output, reaching a point just one per cent short of the record set in 1987. The railways attributed the latest progress to several factors: the strong performance of the North American economy, a double-digit increase in Canadian exports (which were helped by a weak Canadian dollar) and the growth in double-stack container services.

Output prices were relatively stable in the late 1980s and early 1990s, with a major drop (7.3 per cent) in 1994. Although market conditions enabled railways to increase freight prices of some commodities by 1.5 per cent in 1995, price levels were still six per cent lower than in 1986.

Since 1984, freight revenues have increased only 4.5 per cent, whereas the volume of activity actually rose 6.5 per cent. The decline of nominal rail freight prices since 1987 means that the cost of rail freight has dropped by almost a third in real terms since 1981. This represents a transport cost reduction of close to $2 billion for shippers and producers in Canada.

Freight revenues of the trucking industry grew from $4.6 billion in 1981 to $11 billion in 1995, a robust annual growth of 6.3 per cent. The slowing of revenue growth between 1989 and 1993 to 1.3 per cent per year was due mainly to flat prices. However, the growth of trucking revenues rebounded dramatically ^{note 7} (25 per cent) between 1993 and 1995.

In relation to the business sector of the Canadian economy, trucking prices have fallen 19 per cent in real terms over the period studied. The decline in relative pricing of trucking was much steeper in the second half of the 1980s (three per cent per year) than in the first half, when trucking prices paralleled those of the general economy.

In terms of real output, freight trucking output registered a steady growth of 4.4 per cent in the first period (1981 to 1986) and 6.2 per cent in the second period (1986 to 1995). As Figure 15-8 indicates, the performance of the trucking industry in general seems to follow the cycles of the economy. As much as 97 per cent of the variations in trucking output are explained by variations in the economy. There is also evidence of strong demand elasticity from the economy, the result of both modal substitution from rail and private trucking and the increase in transport intensity in the economy.

All Carriers

This section summarizes the analyses for all transport modes. Figure 15-9 shows annual changes in output and price for the transport sector and for the economy in general. In 1995, the larger transport firms generated $26 billion from the inter-city movement of goods and people. Note that the revenues of the transport industries have remained stable around the $23 billion mark between 1988 and 1993. Revenues surged 10 and five per cent in 1994 and 1995, respectively. Per annum revenue growth was four per cent, divided between price changes (1.8 per cent) and output changes (2.2 per cent).

The transport sector has generally trailed the economy in both output and price, a trend that has been especially evident since the late 1980s. Lower transport prices have meant savings for private and business transport users exceeding $3.5 billion in 1995. Savings were particularly large for freight users, representing close to 24 per cent of the actual cost borne by shippers in 1995. For passenger services, however, price increases above the inflation trends in the economy represented an additional $750 million in user costs.

Efficiency of Transport Industries

The preceding section has demonstrated that users of transport services have benefited from a reduction of transport prices since 1981. Lower prices can be sustainable beyond year-to-year volatility only if productivity increases and unit costs of production fall. If lower prices are not coupled with a reduction of unit costs, the financial performance of the firms is in serious jeopardy.

Productivity is a term used to express, in terms of physical outputs compared to physical inputs, how efficiently resources are used; it is equal to the ratio of output over input, as outlined earlier. Unit costs express this relationship in dollar terms. All costs are factored in, including a notional return on capital, to reflect the cost of all the resources used in the production of transport services. However, nothing guarantees that the firms' returns will be sufficient to ensure the repayment of the debt and provide a sufficient return to the shareholders. In this analysis, actual operating margins of the firms are replaced by their "opportunity" cost of capital (COC).

A COC is developed for the capital goods - such as property, plant and equipment - used in each transport sector, measured by their net book value. The method of financing - equity, debt or deferred taxes - was determined from the firms' capital structure. A blended cost of capital rate was then estimated, which, multiplied by fixed assets, yielded the COC of the firms. The COC was substituted for the operating margin of the firms to determine their fully allocated cost, and the share of each factor of production was computed from total costs, not total revenue.

Cost Structure of Transport Firms

Details of the cost structure of the firms in 1994 and over the 1981-1994 period are included in the data compendium.

Labour makes up, on average, one-third of the costs of transport industries, the highest share being recorded by the bus industry (41 per cent), the lowest by the marine sector (25 per cent). Labour's share of costs in the rail mode averaged 42 per cent for the period but dropped to 38 per cent in 1994. Fuel costs are in decline for all transport industries, particularly the air industry, which climbed to a high of 27 per cent in 1982 and fell to 13 per cent in 1994.

Other cost elements are more specific to each mode. The trucking industry, for example, has a high (27 per cent), and increasing, share of owner- operator cost. The air industry has relatively high marketing costs - 14 per cent of industry costs in 1994.

The last three cost elements in particular - leasing, depreciation and a notional return on capital invested - reflect the full cost of capital. Together these three cost elements are relatively fixed over time, accounting for 18-19 per cent of all transport firms' costs. However, there are variations among modes and within the cost of capital categories. The cost of capital is highest in rail, reaching 29 per cent in 1994, and lowest in marine transportation (13.4 per cent in 1994). The marine sector's lower cost share reflects its use of largely depreciated assets and the access to vessel chartering, which can be compared to leasing.

Increasingly, transport firms are relying on leasing as a source of capital (see Figure 15-10). Nowhere is this trend more pronounced than in the air industry, where some 40 per cent of capital costs are incurred from leasing. On average, the proportion of the cost of capital arising from leases across all modes now exceeds 25 per cent.

Bus and Marine Transport Industries

The intercity bus industry's productivity increased by 2.6 per cent per year in the early 1980s. However, these efficiency gains were overshadowed by increases in input prices, especially for labour. Efficiency in the bus sector declined over the next two periods: 1984 to 1988 and 1988 to 1991. In recent years, however, productivity rebounded and the industry has been able to recover the losses experienced in the middle and late 1980s (see Table 15-1). In spite of this improvement, this industry registered the largest unit cost increases of all transport modes.

Until the early 1990s, the marine transport industry was able to adjust its resource utilization to declining demand. Productivity increased an average 2.5 per cent per year while unit costs rose only 1.5 per cent per year. However, a productivity decline of 11 per cent in 1992 had the effect of offsetting a large part of the earlier gains. Overall productivity gains (1.3 per cent per year) were much slower than increases (3.5 per cent per year) in the prices of the resources used by marine carriers. As a result, industry unit costs climbed 32 per cent between 1981 and 1994.

Airline Industry

Total factor productivity of the airline industry in 1995 was about the same as in 1986, and nine per cent above 1981 levels. The trends are not continuous; four periods can be distinguished. In the first - 1981 to 1984 - productivity increased by a robust 4.3 per cent per annum. This is followed by two periods of productivity decline, minor in 1984 to 1988, but a serious 14 per cent drop between 1988 and 1991. Since 1991, productivity has been increasing at a rate of 3.7 per cent per year (see Table 15-2).

These results must be interpreted in the context of major changes in the industry: two recessions and the process of deregulation. Not only did the firms have to consolidate individual companies, and develop networks of regional affiliates, they also had to rationalize their business segments. For obvious reasons, this has had an impact on efficiency.

A major factor in the performance of the airline industry is the uneven nature of any fleet renewal program. Typically, when an airline renews its fleet, it does not replace the aircraft gradually, but in multiples, so as to benefit from economies of scale in maintenance and training. The challenges were compounded when the major fleet renewal of the Canadian airlines coincided with the recession in the late 1980s and early 1990s. These factors had the effect of artificially increasing the capital output ratio of the industry, thus decreasing capital productivity.

Figure 15-11 illustrates changes in capital and variable (i.e., all non-capital costs) productivity.

Since 1981, although capital productivity has fallen by almost two per cent, variable productivity has been increasing at a yearly rate of close to one per cent. In particular, labour productivity showed an annual 2.4 per cent growth. Gains were particularly strong in the 1991-1994 period. The airline industry's performance in labour productivity and unit cost was superior to the performance of the economy in general. Offsetting the labour productivity gains were declines in the product-ivity of capital and other materials and services, which reflect the increasing reliance on outsourcing in the industry.

The input price index of the airline industry grew 3.1 per cent per annum. This was higher than the inflation rate in the general economy, which rose by 2.6 per cent per year. Reductions in the price of fuel were not sufficient to moderate the rising cost of other inputs, such as labour, capital and other materials and services. Increases in unit cost of all factors of production averaged 2.5 per cent annually. Until the mid-1980s, the unit cost performance of the airline industry surpassed all the other transport industries. After 1987, however, the rapidly rising unit cost of the airline industry has been in contrast to the relative cost stability of the other major transport industries: truck and rail.

Rail Freight Industry

Total factor productivity of the railway industry improved by 50 per cent over the 1981-1995 period, with a 13 per cent increase over the last two years (see Table 15-3).

While CN and CP Rail had about the same total productivity growth over the past fourteen years, their major gains occurred in different periods. CP showed strong productivity growth in the 1981 to 1986 period, then slowed until 1992. Conversely, CN's major productivity gains were achieved after 1986, when a major transfer of activities to Via Rail took place. Both railways achieved significant productivity gains in the last three years (1993-1995).

The most important source of total productivity gain in Canadian railways is the increase in labour productivity, which has doubled since 1981. Between 1992 and 1995, labour productivity increased by 28 per cent, as a result of workforce downsizing and strong output growth. Fuel efficiency is also a contributor to overall productivity gains.

High labour productivity did not result in lower unit costs until 1994 because of large wage increases in the early 1980s. Average labour compensation rose 8.7 per cent per year between 1981 and 1984, producing a significant rise in unit labour costs. Unit labour costs have since started a downward trend, which has accelerated in recent years.

In spite of good productivity performance, declines in total unit costs of the rail industry began only after 1992. Because input prices grew faster than productivity, the cost per unit of output increased at an average annual rate of 1.8 per cent between 1981 and 1992.

Trucking Industry

Productivity in the trucking industry increased by 31 per cent between 1981 and 1994 (2.1 per cent per year), rising eight per cent in 1994 after stalling for three years. Productivity gains were more pronounced during the second half of the decade, after the deregulation of the industry (see Table 15-4).

Before 1990, the largest partial productivity gains (4.8 per cent per annum) were recorded by labour, partly because of the substitution of owner-operators for internal resources. These gains slowed considerably and could only be described as minimal in the early 1990s, then jumped in 1993/94. However, part of the large increase in 1994 may be due to changes in survey methodology. ^{note 8}

Input prices increased by 3.1 per cent per year between 1981 and 1994, comparing well with general inflation levels, which averaged two per cent in the same period. Prices of purchased transportation services, capital goods and fuel either dropped or showed minimal increases.

Per unit of output, the performance of the trucking industry is note- worthy, showing a 0.7 per cent per year drop in labour unit costs between 1981 and 1994. By comparison, unit labour costs in the economy increased 2.4 per cent per year over the same period. Overall, unit trucking costs advanced one per cent on a yearly basis, a performance surpassed only by that of rail freight carriers. With output prices keeping pace with unit costs, the financial position of the industry has remained virtually unchanged.

Trucking Productivity Trends by Region

Results were tested by region, based on the carriers' domicile: Atlantic Canada, Quebec, Ontario, the Prairie Provinces and British Columbia (which includes the Territories). The results must be regarded with some degree of caution. Price and output indices, key factors in productivity measurement, were developed on the basis of traffic to and from the regions, but the traffic composition does not necessarily coincide with that of carriers domiciled in each region.

The strongest output growth, by far, was recorded on a regional basis by Atlantic firms, more than doubling that of Canada as a whole. Quebec, Ontario and Prairie firms had growth rates similar to the Canadian average for the whole period, while Ontario lagged behind the rest of the country in the second half of the 1980s. British Columbia firms, though out-paced by the rest of the country in the first half of the 1980s, later rebounded to show a stronger performance in the latter years.

The divergence in productivity among the regional components ( Figure 15-12) of the industry is not pronounced in spite of significant divergence in output growth. This confirms the absence of economies of scale in the industry, and is especially evident when one compares the productivity of the Quebec and British Columbia carriers, which have the smallest size but show the strongest productivity performance.

Total Transport

Total productivity in the transport sector in Canada grew by 1.8 per cent over the 1981-1994 period. The transport sector outperformed the economy in general in the first half of the 1980s. Productivity growth slowed somewhat between 1986 and 1991 in transport but remains positive, whereas producti-vity in the general economy has fallen. Since the end of the recession in 1991, the productivity gains of the transport sector have been robust, at more than 3.5 per cent per year (see Table 15-5).

Since the mid-1980s, productivity gains of freight carriers have been accelerating, while passenger carriers have reported mixed performance. The productivity of the latter dropped considerably in the second half of the decade but bounced back afterwards. Annual changes in productivity are depicted for selected modes in Figure 15-13.

Unit costs are influenced by the behaviour of productivity and factor prices. Transport sector factor prices increased at an annual rate of 3.3 per cent between 1981 and 1994, exceeding the rate of inflation in the Canadian economy during the same period (3.0 per cent). Wage gains in transport outpaced gains made elsewhere, but fuel prices declined and capital prices were moderate. Total unit cost of the transport sector advanced about 3.1 per cent in the first half of the 1980s, matching the overall pace in the Canadian economy. The pace slowed, except for passenger carriers, between 1986 and 1991, when unit costs increased by 6.3 per cent per year. Unit costs have been in decline since 1991 for both passenger and freight carriers.

Endnotes

 

1. For practical reasons, the assessment is confined in this first annual review to firms that receive the majority of their revenues from transport: railroads, airlines, trucking companies, shipping lines, and bus companies. It does not include the enterprises providing infrastructure or services that either still are or were until recently government-owned and operated - in particular the airports, air navigation services, ports, or the Seaway. As information on their finances and operations equivalent to that for carriers becomes available, the assessment of performance will be extended to these other categories.

    

2. Input price and unit cost are two different notions. The first refers to the price of input per unit of utilization. The second refers to the price of input per unit of production, calculated by dividing input price changes by productivity changes.

    

3. The technique uses the T”rnquist approximation to the Divisia Index. See D.W. Caves, L.R. Christensen and W. E. Diewert, "Multilateral Comparisons of Output, Input and Productivity Using Superlative Index Numbers," Economic Journal, March 1982.

    

4. Aggregation and Accuracy in Measuring Total Factor Productivity, Evidence from Rail Productivity Studies, W.W. Waters and M. Tretheway. Annual Proceeding of Transportation Research Forum, 1995.

    

5. 1995 data for the marine mode are estimates.

    

6. Data for shipping industry in 1995 are estimates.

    

7. Changes in survey methodology may explain the strong growth of 1994.

    

8. Until 1993, an annual survey (the Annual Motor Carriers of Freight Survey) covered all trucking firms earning at least $1 million. Since 1994, the trucking industry is covered by a two-stage survey on a sample basis.

Appendix

Scope of Analyses and Statistical Sources

Scope of Analyses

The productivity analysis of the airline industry covers the activities of Air Canada and Canadian Airlines International (CAI) and their main affiliates or partners. Air carriers associated with Air Canada included: AirBC, Air Nova, Air Alliance, Air Ontario and Austin Airways. Those associated with CAI included CP Air, Pacific Western Airlines, Eastern Provincial, Nordair, Nordair Metro, Quebecair, Wardair, Time Air, Air Atlantic, Inter-Canadian, Canadian Partner, and Quebec Aviation Ltee. The analysis of price and output measurements is expanded to include the main charter operators since 1988. As much as 98 per cent of the activities of Levels I and II are covered. When all airlines are included, the scope of analysis is reduced to 86 per cent.

In the bus industry, only the intercity scheduled and charter bus services (interurban as well as rural bus) are reviewed; carriers whose main activities are transit or school bus operations are excluded. In 1994, the revenues of intercity carriers represented 12 per cent of all bus carriers.

The analysis of the performance of for-hire trucking focuses on firms with sales equal to or greater than $1 million before 1989. Since that year, all trucking firms covered by the sample frame of the Motor Carrier Survey of Statistics Canada are included. This ensures that similar firms are analysed over time. Smaller firms (15 per cent of industry revenues) were excluded from the analysis because of data inconsistencies. Individual carriers whose main activity is the movement of household goods (five per cent of industry revenues) were also excluded.

In rail, the analysis covers Canadian Class I freight railways, CN and CP. Together, they generate about 90 per cent of rail freight operating revenues in Canada, the rest being accounted for by regional railways and terminal/switching operators. Both CN and CP own and operate several rail lines in the United States, but only their Canadian operations are covered. The analysis of price and output measurements is expanded to include the activities of Via Rail.

In marine transportation, the analysis is confined to Canadian-domiciled for-hire carriers using both Canadian and foreign flag ships. It does not include foreign-based shipping firms that provide marine services to Canadian exporters and importers. The analysis of price and output measurements is expanded to include the passenger activities of government ferry services.

In aggregate, the universe of the price and productivity analyses reflects 97 per cent of larger, intercity and non-government carriers. When all carriers are included, the share is reduced to 73 per cent. More detail is provided in Table 15-1A.

Data Sources

In the airline industry, the sources of output data were Statistics Canada's Air Carrier Statement 10, entitled "Scheduled Services, Revenue Operating Statistics," and Air Carrier Statement 12, entitled "Charter Services, Revenue Operating Statistics." The source of cost data was a special run from Statistics Canada on Air Carriers, which included: Air Carrier Statements 20 ("Balance Sheet"), 21 ("Income and Expenses"), and 32 ("Fuel and Wages"). The main source of data for the bus industry was the Statistics Canada publication Passenger Bus and Urban Transit Statistics, (Cat. 53-215). Special tabulations were also obtained from the Transport Division of Statistics Canada. Output prices of the Canadian trucking industry have been derived from Statistics Canada's For-Hire Trucking Commodity Origin/ Destination Survey files. Cost data were derived from the files of Statistics Canada's Motor Carrier Survey (Cat. 53-222) at the individual firm level up to 1993. After that year, special runs by region of domicile of the carriers were prepared by Statistics Canada.

The main data source for the rail industry was a UBC study on the productivity of CN and CP, 9 the Statistics Canada files of Rail Transport (Cat. 52-216 ), Canadian Transportation Agency files, annual reports of CN and CP, and the Canadian Institute of Guided Ground Transport (The Impact of Regulatory Reform on the Canadian Railway Industry, November 1992).

The main data source for the shipping industry was Statistics Canada - files of the water transport industry survey of for-hire, government and private carriers (Cat. 54-205). Other sources, such as the Canadian Shipowners Association, were also used.

Economic indicators were derived from Statistics Canada's national accounts.

Additional Information

The source data for the tables and graphs in Chapter 15 can be found in the data compendium of this report. Working papers on the methodology, and detailed results for some industries are available from the Economic Analysis Directorate of Transport Canada.