A FORECAST OF SCHEDULED AIRLINE
REVENUE PASSENGER TRAFFIC IN BRAZIL
by
Anthony Eric Rickards, B.A.
A thesis submitted to the Faculty of Graduate Studies and Research in partial fulfilment of the requirements for the degree of Master of Arts.
Department of Economies and Political Science, McGill University, Montreal. March 1961 A FORECAST OF SCHEDULED AIRLINE REVENUE PASSENGER TRAFFIC IN BRAZIL
TABLE OF CONTENTS
Page No.
Chapter I Introduction •••••••••••••••• 1
Chapter II An Analysis of the Brazilian Economy ••••••••• 3
Chapter III The Brazilian Transportation System • • • • • • • • • • • • . • • • • • • • 44
Chapter IV An Analysis of the Economie Position of the Brazilian Airline Industry •••••••••• 52 Chapter V Discussion of Methods used in Forecasting Air Passenger Traffic ••••••••••••••••••• 66 Chapter VI Forecast of Revenue Passenger Kilometres to be performed by Scheduled Airlines in Brazil ••••••••..•••••••••• 116
Appendix I to Chapter VI Calculations of the coefficients of correlation between indices of industrial production and numbers of air passenger kilometres performed ••••••••••••••••• 129
Appendix II to Chapter VI Estimate of the number of private automobile inter city passenger-kilometres performed in Brazil in 19 57 . • • • • • • • • • • • • • • • • • • 15 3 Appendix III to Chapter VI Calculation of the coefficient of correlation between Brazilian Gross National Product and the Index of Industrial Production •••••••••••••••• 157 Bibliography ...... 163 CHAPTER I
I N T R 0 D U C T I 0 N 1
INTRODUCTION
CHAPTER I The objective of this thesis is to forecast the scheduled airline revenue passenger traffic which will be carried by Brazilian airlines during the period
1959 to 1965. In order to provide a firm base for this forecast the following analyses have been made:
1) an analysis of the economie
position of Brazil.
2) a descriptive examination of
the Brazilian transportation
system.
3) an analysis of the economie position
of the Brazilian airline industry.
4) an analysis and discussion of
methods used in forecasting air
passenger traffic.
5) an examination of past air traffic
statistics.
The analysis of the Brazilian economy has been undertaken in order to provide a basic knowledge of that country 1 s major economie characteristics and thus to provide knowledge of the current economie trends which are expected to affect the field of transportation. r 2
Any forecast of the future performance of
an industry must naturally take into account the
specifie conditions existing in that industry and
therefore an analysis of the Brazilian airline industry
is included in this thesis.
The objective of this paper being to forecast
airline passenger traffic, it is considered essential
that an analysis and discussion of forecasting methods
be included, in order that the particular forecasting
method used here be viewed in a larger framework.
Having taken into consideration and been
guided by the conclusions drawn from these analyses,
a traffic forecast has then been formulated. CHAPTER II
AN ANALYSIS OF THE BRAZILIAN ECONOMY 3
AN ANALYSIS OF THE BRAZILIAN ECONOMY
CHAPTER II
Before embarking on a study of Brazil 1 s economy proper it would be well to give a very brief description of its main geographie, climatic and population charac teristics.
GEOGRAPHY AND CLIMATE
The United States of Brazil, with an area of
3,287,195 square miles (3,270,233 of land area and 16,962 of water), occupies about 47 percent of the South American continent and is one of the largest countries of the world. Its land area is only 7 percent less than that of the continental United States of America, including
Alaska. Brazil 2 s greatest north-south and east-west dimensions at the points of their greatest extent are almost identical - 2,684 and 2,689 miles, respectively.
Its boundaries extend for 14,370 miles, of which 4,603 miles is a coastline along the Atlantic Ocean. Brazil borde rs on a11 other countries of the continent exce pt
Chile and Ecuador.
The greatest part of Brazil is made up of highlands, whi ch have no great peaks. North of Salvador,
Bahia, the rise from coast to the interior is graduai, but from Salvador south to Porto Alegre, Rio Grande do
Sul , there is a sharp rise known as the Gre at Esca rpment, which so r esemble s a mountain r ange tha t one part o f it is called the Serra de Mar. 4
About 57 percent of the land is in plateaux, varying from approximately 650 to 3,000 feet in elevation;
40 percent consists of lowlands with an altitude of
less than 650 feet and 3 percent exceeds approximately
3,000 feet in altitude, with the highest peak reaching
9,482 feet above sea level.
Because of the great north-south extension
(50° 16' 19" N. to 33° 45 1 09" s.) and the varying altitudes, there is a considerable range of climatic
conditions. The only section lying in the Temperate
Zone is the area to the south of the city of Sao Paulo, which is on the Tropic of Capricorn. However, the
climate of large areas of central and northeastern
Brazil is moderated by the altitude or by sea breezes.
In the tropical and sub-tropical sections
of the country the seasons are marked more by wet and dry periods than by changes in temperature. In them,
summer is generally the wet season and winter the dry,
but this is not uniformly true.
Brazil is free from cyclones, hurricanes,
and similar disturbances. ·
POPULATION
The total population was 51,976,397 in the census of 1950, a rise of 26 percent from the 41,114,315 recorded in the census of 1940. The population is estimated officially to have increased to 62.7 million as of July 1, 1958. All but a small proportion of 5 the growth is the result of natural increase, as immigration into the country (not deducting for
\ \ emigration) was only 114,085 in the years 1940-49, and 491,510 in 1950-57. The census recorded 98 percent of the population as Brazilian-born.
The greater part of the population in the past was in the north-east section of the country but over the years the major concentration of the population has shifted southward. The population of the Rio de Janeiro-Sao Paulo area has been
augmented considerably by migration within the
country and a population movement to the westward is occurring. The constitution provides for the
transfer of the Federal Capital from Rio de Janeiro to the central plateau; a new capital city, Brasilia is under construction and nearing completion in
Goias, with the move to the new capital officially having taken place in April 1960. The emphasis
on the development of Brasilia is furthering the development of the central region.
GOVERNMENT ECONOMIC POLICY
The policies followed by the Government
since the 1930 1 s have encouraged the development of industry. Import controls and, since August 1957, a highly protective tariff have operated to protect domestic industry and, in some cases, to reserve the entire market for domestic production. The shortage of foreign exchange over much of the period and the 6
consequent reduction in imports of 11 non-essentials 11 acted as a further spur to the development of consumer goods industries. The exchange control system in effect since late 1953 grants relatively favourable treatment to raw materials and industrial equipment considered essential to the economy, by making foreign currency available for their purchase at a special, low exchange rate.
President Juscelino Kubitschek, who was inaugurated on January 31, 1956~ announced production plans, with goals to be attained during his 5-year term of office. The emphasis is on power, transportation, and agriculture, and on industrial education, but production programs also were established for basic industrial products. To implement these plans, some federal agencies of a highly specialised character have been established. The Development Council, created in February 1956 to centralize and coordinate plans for economie development, functions principally through working groups of technicians appointed to analyze specifie problems and to make recommendations for appropriate action.
The principal financing agency for the economie development program is the National Economie Development
Bank, which was established in 1952 to handle financial operations of projects that were designed mainly for the rehabilitation and reequipment of transportation 7 facilities and for the development of power resources recommended by the Joint Brazil-United States Economie
Development Commission. The bank originally derived funds principally from forced loans assessed as a percentage of income tax payments, but it now acts as the agent of the Government in handling receipts from special taxes and other funds to be used for specifie purposes, such as road construction and rural electri fication. The bank also handles the cruzeiro loan funds resulting from the sale of surplus United States agricultural products to Brazil.
Manufacturing companies in Brazil are, for the most part, privately owned. Exceptions are the country 1 s largest steel plant, The National Steel
Company at Volta Redonda; The National Alkali Company, which is not yet in full operation; and The National
Motors Factory, established to manufacture aircraft motors and now utilized for the production of motor trucks and tractors. These are stock companies of mixed Government and private ownership, with Government stock control. The Cubatao refinery of the Government petroleum monopoly, Petrobras, is the centre of a group of associated plants producing petrochemicals.
In the field of agriculture, the government has initiated programs to promote agricultural development which include the expansion of agricultural credit, financing and priee support for a number of crops, facilitation of imports of agricultural machinery, and 8
warehouse const~~ction. Special encouragement is given to wheat growing through the maintenance of high priee
supports. Research work is carried on by both the Federal
and State governments.
STRUCTURE OF THE ECONOMY
There follows here a description of the structure
of the Brazilian economy. Except where otherwise indicated,
the source for statistics quoted in this section is the
1959 edition of the Anuario Estatistico do Brasil,
published by the Conselho Nacional de Estatistica, in Rio de Janeiro.
Manufacturing
Between 1944 and 1955 the index of physical industrial production rose from 84 to 163 (using 1948
as the base year, with an index of lOO). While this rise is in itself significant, what is even more
significant is that industry experienced great diver sification during this period. Iron, steel, cement and rubber output increased sharply (iron and steel - sixfold;
cement - threefold; rubber- threefold), whereas industries auch as textiles and food (both long established) only increased by about a half. The table on the following page shows the indices of industrial production for the period 1944 to 1957. 9 TABLE I
INDICES OF VOLUME OF PHYSICAL INDUSTRIAL PRODUCTION 1944 - 1957 (1948 MONTHLY AVERAGE= 100)
: SUMMARY PRODUCTION GOODS CONSUMPTION GOODS
~ ~15 ~ 0 +)C) U1 0 ::l ::l CH U1 ~-~(") ·ri +) fl.l bO (l)'tj U1 CH (1) 0'-' C) +) "" ~C) ~ C) 0 (1) ::l b.O 0 \ p ~ ,....., C) ~ b.O +) M f""'4 +) Cil C) M ::l fl.l ::l co ~ co ·a +) M (1) ~J: co co M C) ""+)M · co 'tj'tj CO'tj co Cil ""f""'4- «>- .o- +)- «>- 'tj.-\ «>- co- ~~ C) (1) ,j.J · 0 0 ~ 0 ~-~r-I ~-;:; Cil M PiC'! ,.QC'\1 ~~ ""~C'\1 OC'\1 Or-1 l>r-1 .Or-1 (1) ~ ' 0 j M 0 0 0 r-Ia> Cil'-' ::l'-' ..C:'-' 0'-' 0'-' r-IO Œ>'-' J~ p.. ~ - ~ Œ>'-' «>'-' Year ' E-l l=l-tc.!:l UC!) ~- u C!)U P::: E-1 Cf). ~ r:Q E-1 e~ ÇJ:Il=l-t
1944 84 64 93 90 73 45 75 55 64 103 91 101 69 52 70 69 1945 86 68 93 91 70 81 43 76 79 83 96 98 98 82 55 83 76 1946 93 79 97 87 74 94 64 84 88 108 102 101 88 94 77 105 83 1947 93 86 94 89 82 95 79 91 86 101 95 83 93 93 94 lll 90 1948 lOO 100 100 lOO 100 100 100 100 100 lOO lOO 100 100 lOO 100 100 100 1949 106 116 106 104 115 113 119 116 123 105 107 101 106 102 116 87 108 1950 118 135 117 104 125 129 153 133 140 123 114 112 118 120 133 92 117 1951 129 150 124 118 130 154 167 140 159 153 110 135 128 143 156 113 126 1952 137 157 130 120 145 161 164 140 172 143 116 124 137 l55 172 149 136 1953 146 177 137 120 183 190 197 156 194 147 122 128 147 161 178 151 138 1954 156 193 147 126 217 186 229 161 193 158 144 122 139 165 205 144 154 1955 163 198 156 142 234 196 233 168 199 13l 151 147 157 175 224 132 168 1956 • • • ... . • • • 173 294 ••• 265 • •• 186 • • • • • • • •• • • • • • • • •• 133 188 1957 • • • • • • • • • • •• 287 • • • 283 • • • 196 • •• • • • ••• .. . • •• ••• 138 206
SOURCE: - Fundacao Getulio Vargas (1) Weighted indices adjusted for annua1 and seasonal variations (2) Simple indices adjusted for annua1 and seasonal variations
(3) Simple indices 10
The trend to diversification resulted in a
greatly increased production of non-consumers 1 goods,
compared to the increase in production of consumera'
goods. This growth in the non-consumers' goods industries,
however, was also the result of the development of
supporting industries - industries producing machinery
of all types, transportation equipment, electrical
equipment, etc.
Automotive Eguipment
In the field of automotive equipment production
particularly, there have been significant advances. In
1955 there was no automotive industry, other than several
assembly plants, and little incentive existed for manu
facturers of automotive parts. In 1956 a governmental
industrial planning group was established to encourage
development in this field.
By 1958, 9 automotive vehicle factories and
about 1,200 parts factories were in operation, of which
at least 150 operate on a large industrial scale.
Production of motor vehicles in 1958 totaled
62,707 units, as compared with 33,000 in 1957 and 6,087
in 1956. 1 Units produced included the following:
Trucks 35,369 (National Motor Factory -
F.N.M. 3,800; Ford, 10,847; General
Motors, 9,353; International Harvester,
450; Mercedes Benz, 10,693; DKW -
Vemag, 226)
Buses 361, (Merce d es Benz) 1 u.s. Department _of Commerce, Economie Developments 11
Jeeps 14~216~ (Willys~ 13~163;
DKW- Vemag, 1,053)
Utility 7,642, (Volkswagen, 4~819;
Willys~ 2823)
Passenger cars 5,119, (Romi-Isetta~
1~400; DKW - Vemag, 3~719 of which
~ 1,599 were station wagons).
The number of vehicles assembled
dropped to 183, compared with 5,012
in 1957. Brazil's first passenger car, the first four-
door sedan made in Latin America, was manufactured by
DKW - Vemag during the year. This company also produced
its first trucks. Volkswagen passenger car production
started in January 1959, at a daily rate of 20. Willys Overland, with $3.5 million investments from the United
States, was already expanding its original operations
and, although plans for the production of Plymouth -
Savoy automobiles were changed to the production of
five passenger Aero Willys (produced in the United States
until 1953), to start in 1959, Willys made an agreement
with Renault of France to manufacture Dauphine passenger
automobile s, also starting in 1959.
Imports of vehicles, including chassis with
engines for trucks and buses, increased to 74,480 from
42,394 in 1957. The largest increase, 130 percent, was
recorded for imports from Germany (23,733 units), although
the largest supplier remained the United States (46,180 units). 12
The production of trolley buses, as well as
roadrollers, was also inaugurated in 1958. A firm was
formed to manufacture materials-handling equipment-
crawlers, excavators, tractor shovels, and the like.
The expansion of the automobile parts industry
was an outstanding aspect of the growth in Brazil 1 s
economy in 1958, about 297 firms having been inaugurated.
Steel
In 1956 Brazilian steel mills produced 1.36
million tons of steel ingots and in 1958 production ~ reached 1.6 million tons. However, in spite of this
increase, the expected surplus of domestic production
over domestic consumption has not materialised.
Consumption of ivon and steel during the
period 1956-65 was underestimated in the plans, which
expected surplus production after 1960. As it now appears,
the deficit in domestic production will increase from
600,000 tons in 1958 to over 2 million tons, by 1965,
even if planned mills are completed. This situation
is an effect of the increasing industrial tempo,
especially the stimulation of the automotive vehicle
and parts sector, as well as the new shipbuilding sector.
In order to increase production, two large
new producers are being established - Usiminas, in the
State of Minas Gerais, and the Cia Siderurgica Paulista
{COSIPA) in Sao Paulo. The first, which is being developed
in association with Japanese capital and experience,
will be ready after 1960 and will produce 500,000 tons
2 Ibid, P. 5 13 of steel ingots in its first stage, being later developed to one million tons. It will cost 12,000 million cruzeiros and part of its capital has been subscribed by the State of Minas Gerais, the National
Steel Mill, the Rio Doce Valley Company and the
National Bank for Economie Development.
The building site for the new 50,000 ton
COSIPA steel mill is being prepared at Piacaguera,
Sao Paulo, as a result of financing having been assured from the Federal and State Governments, as well as from the United States, England, and Italy, to cover the costs of imported materials and equipment.
A new mill in the State of Santa Catarina utilizing national coal in the production of average and light weight shapes, was proposed to the Brazilian Congress by the President of Brazil.
Steel manufactured consumer durables produced in Brazil include such items as refrigerators (production of 210,000 domestic refrigerators is planned for 1960), washing machines (30,000 in 1957), sewing machines, vacuum cleaners, floor polishers, typewriters, motor cycles and bicycles.
Chemicals
The chemical and pharmaceutical industry has become a major factor in Brazil's economy and, aside from the automotive sector, seems to hold the most promise for the future. Although substantial 14
quantities of basic materials must still be imported,
self-sufficiency is expected within a decade.
The petrochemical industry has developed
recently, with investments of about $75 million, of
which about $60 million are from foreign sources.
This industry has been centred around the new refinery
at Cubatao, and is expected to develop at the refinery
being built at Duque de Caixas.
Brazil's infant plastic materials manufacturing
industry is entering a period of growth. At present
both domestic production and importa are quickly
absorbed, and the market is constantly growing at an
annual rate of about 15 percent.
In March 1958, a new pharmaceutical plant,
one of the most modern and well equipped, was inaugurated
by Merck, Sharp, and Dohme near Campinas, Sao Paulo.
It started production of vitamins, mineral supplements,
and hormone ster~ids. The newest equipment to manufacture
compound tablets has been installed. Construction of
what is intended to be the l a rgest and mo s t modern of
pharmaceutical plants in Latin America was started by
the Brazilian subsidiary of the Swiss firm, Ciba. A
Sao Paulo chemical firm and an Italian company have
combined to produce raw materials for the pharmaceutical industry.
The tota l production of fertilize rs {of which
phos ph atic f e rti l i zers are t he gre ate r p a rt) which was
249,000 tons in 1957, has been increasing at approximately
17 p e rcent annua lly, but is unable to meet the growing d emand. It i s estimated tha t the output o f organi c 15
fertilizers in 1958 was between 30,000 and 35,000 tonsY
Nitrogenous fertilizer production is also increasing.
Other industrial expansion
Great expansion is also taking place in the
pulp and paper industry and in the heavy electrical
equipment industry, whilst construction of shipyards,
aimed at the establishment of a shipbuilding industry,
is now under way.
AGRICULTURE
Real output of Brazilian agriculture has,
since 1948, outstripped the growth in population.
There has also been a significant trend in Brazilian
agriculture, this trend being signified by the
difference in the rates of growth of crops produced
for export and crops produced for domestic consumption.
Production of domestic food crops has been growing
steadily since the war, whilst production of export
crops has remained at the same level or decreased.
To illustrate the situation with regard to export crops - in the crop year 1933~34, 29.6 million bags of coffee we re produced. The average production
1evel for the past few years has been between 17 and
22 million bags. In 1955 and 1956 production of
3 Ibid, P. 6 16 cocoa decreased, remained steady ataround 167,000 metric tons in 1957, and increased slightly in
1958 and 1959 to around 172,000 and 175,000 tons respectively. The 1958 production of cotton was estLmated at 289,000 net metric tons of lint, compared with 290,000 tons for the preceding year and 366,000 tons for 1956.
With regard to domestic food crops, wheat stands out as that crop which has shawn the most dramatic growth in recent years. In 1956 the government fixed the priee of wheat at 460 cruzeiros per 60 kilogram bag, and, with this support, production in 1956 reached 1.2 million metric tons. However, as a result of bad weather, the 1957-58 crop year production fell to only
300,000 tons. In spite of this setback, due to abnormal weather conditions, Brazil saon hopes to be completely self-sufficient in wheat, and thus avoid her previous heavy expenditure of foreign currency for this commodity.
There have also been great expansions in the production of potatoes and maize, while Brazil's production of s ugar has grown to be the third largest in the world, after Cuba and Russia. 17
The two major factors accounting for the substantial increase in foodstuff production are increased productivity and acreage. Greater productivity may be traced to three factors: use of new areas with more fertile soils, mechanization, and improved agricultural practices. The opening up of fertile agricultural land in western Sao Paulo, northern
Parana and southern Mato Grosso and Goias, brought into production areas where yields are rouch higher than in many of the older agricultural regions.
But improved farm methods, including mechanization, seem to be just as important in accounting for higher productivity as new lands.
Returning to the major export crop - coffee it is necessary to note that this commodity has a very low elasticity of demand, both in terms of income and of priee. Because, however, its production process is long (five years to reach maturity, twelve years to attain high yields), it means that there is no elasticity of production in the short run and a high elasticity in the long run.
Thus, when priees are high, new coffee lands are put under development, but only 5 to 12 years later does an abundance of coffee reach the market. Coffee lands have a productive life of more than thirty years, and so, when coffee priees fall, there is no short run effect on the s upply. 18
However, in regard to this situation, it
must be remembered that Brazil is not the only producer
of coffee in the world and is thus not able to overcome
the difficulties inherent in coffee marketing by
unilaterally controlling the world priee. Any
unilateral action she takes to stabilise priees,
will not be effective, because of the competition
from other coffee-producing countries.
In the face of these conditions attempts
have been made to reach agreement among coffee
producing countries on production quotas.
This agreement, in order to be lOO%
effective, should include all producers, Latin
American, African and others. An exclusively Latin
American agreement will have great impact for sorne
years, but it will also tend to increase production
in Africa, where the European powers are creating
a new type of coffee plantation with a very low
cost structure, based on cheap labour. This is
all the more significant because coffee is still
the sort of product whose quality depends fund
amentally on "manual care" and lends itself to mass picking and handling only at the cost of a reduction
in quality.
However, this problem raises the whole
question of whether it is in Brazil's long-term interest to remain so much dependent upon one export 19
crop (and a labour-intensive crop, at that). At
one time or other in the past Brazil has held a
world monopoly in sugar and in rubber. She lost
both of these and is now in the process of losing
the dominant position she has held in the production
and marketing of coffee. It is possible that, as
Brazilian industrial development continues and as
her domestic market enlarges, she will not need
to rely so heavily upon the revenue derived from
the export of coffee. It is possible, therefore,
that Brazilian coffee production can be more adjusted
to domestic consumption than to foreign consumption.
However, this is the long~term view and
Brazil still has to rely greatly upon coffee as a
means of earning badly needed foreign exchange.
Consequently, the Brazilian government has entered into
agreements in the past and continues to enter agree ments, among Latin American coffee producers, designed
to adjust the coffee supply to meet the demand.
LIVESTOCK AND FORESTRY
Brazil ranks high among the countries of the world in its cattle population, which was
officially estimated as 66.7 million head at the
end of 1956. The hog population, which also is relatively large, was 41.4 million. Other livestock numbers, in millions, were: Sheep, 18.9; goats, 10.3; 20
horses, 7.9; mules, 3.6; asses, 1.9; chickens, 162.4; 4 turkeys, 2.5; ducks, and geese, 5.4.
The area of Brazil's forest land in 1954 is
estimated by the Food and Agriculture Organization
of the United Nations (FAO) to exceed that of any other country of the world except the u.s.s.R., totaling
1,854,000 square miles, or 56.7 percent of the land
area of the country. However, only 463,500 square miles, or 14.2 percent of the land area, is considered
to be accessible forests. Further, only 115,000 square miles, representing 25 percent of accessible forests
and 6 percent of all forest land, is being utilized.
The FAO estimates that 40 percent of the accessible
forest land (185,400 square miles) is privately owned.
The parana pine, which is found principally in Santa Catarina and Parana, is the most important tree from the commercial standpoint, and pine lumber is among the principal exports of Brazil.
The production of wood pulp in 1956 and
1957 was estimated at nearly 300,000 tons annually and considerable additional plant capacity is planned or under construction. Newsprint is produced in reels only by the Klabin plant in Parana, with an annual output in 1956 and 1957 of 38,000 tons and an anticipated production in 1958 of 55,000 tons. The production of other types of paper in 1956 was reported a s 349,000 tons.
4 u.s. Department of Commerce, Basic Data on the Economy of Brazil, P. 6 21
In addition to lumber a number of other
forest products are economically important in
Brazil. The production of the 20 most important forest products was officially valued at 4 billion
cruzeiros in 1956 •
. The collection of rubber, nonelastic gums, Brazil nuts, and a variety of fibers, nuts, essential
oils, and crude drugs is important in the economy
of the Amazon region.
MINERAL RESOURCES
Table 2 below shows the production of the
principal minerais from 1938 to 1956.
TABLE 2 Production of Principal Minerais, 1938, 1950, 1956 (Quantity in metric tons except as indicated)
Product 1918 19'l0 191)6 Iron ore 533,282l 1,987,425 4,085,835 Co al 907,224 1,958,649 2,234,059 Gold-kilograms 4,447 4,082 3,802 Salt 754,871 754,181 798,428 Rock crystal 7472 223 541 Calcined alumina ( 3) ( 3) 17,576 Lead ore (3) (3) 57,958 Scheelite (3) 482 1,305 Manganese 306,025 195,505 310,783 Mar ble 13,176 23,817 39,771 Mica 905 1,813 1,327 1 1939 data 2 Exports 3 Not available Source: Conselho Nacional de Estatistica, Anuario Estatistico do Brasil, editions of 1956 and 1957. 22
The following description of the present state of the mining industry refers only to known mineral deposits. The mineral potentialities dormant in many regions of the vast interior of Brazil are unknown and, evenin the relatively densely settled coastal area, surprising discoveries are sometimes made, reflecting the lack of systematic mineral prospecting in the past.
The iron ore reserves of Brazil are among the richest in the world. A survey of the mineral deposits of east central Minas Gerais, the area where the most important known iron ore reserves of Brazil are concentrated, has suggested that there are probably reserves of 20.7 billion tons of hematite, or 15.2 billion tons of iron.
Export by the Government~controlled
Cia. Yale do Rio Doce, which was organized during
World War II, has been an important factor in bringing about a considerable increase in iron ore production. Exports of iron ore rose from 599,000 tons in 1948 to 1.3 million in 1951 and to 3.6 million in 1957. The Rio Doce company accounted for 3 million tons in 1957. Total production of iron ore in 1957 was about 6.4 million tons.S
5 Ibid, P. 8 23
The Federal Government has established
as one of its goals the promotion of iron ore exports in quantity, and efforts are being made to improve
railway transportation facilities and otherwise
encourage production for export. Expansion plans
of the Rio Doce company provide for the exportation
of 6 million tons of ore annually.
Manganese
Total manganese production in 1957 is
estimated as 811,000 tons. Exports of manganese
from the territory of Amapa in north Brazil began
in 1957 and totalled 668,000 tons for the year.
This project, in which the Bethlehem Steel Company
owns a 49 percent interest, involves deposits
estimated to contain reserves of 30 million tons
of economically usable ore. The Urucum deposits,
near Corumba, Mato Grosso, are reported to contain 6 reserves of 20 million tons of ore.
Copper, Lead, Nickel
Brazil possesses deposits capable of
supplying its nonferrous requirements, but is
presently a net importer of these metals. Imports
of copper in all forms in 1957 totalled 37,000 tons
while production in copper content is estimated at 1,600 to~sJ Lead production has increased in the
6 U.S. Department of Commerce, Economie Developments in Brazil - 1957, P. 5
7 lh!Q, P. 5 24 past few years, and production of pig lead in 1957 is estimated at 6,700 tons, against requirements of 20,000 tons. At least one known large deposit of high grade zinc ore exists, at Vazante, Minas
Gerais. However, no reduction of zinc ore is carried out at present in Brazil and the estimated consumption of about 25,000 tons must be imported.
Imports of nickel in all forms in 1957 amounted to 460 tons and production in nickel content to about 80 tons.
Domestic coal production, which is supplemented by imports, was approximately 2 million tons annually from 1943 through 1954 and increased about 10 percent in 1955 and 1956, declining s1ightly in 1957 to 2,116,000 tons.
Production is localized in the 3 southernmost
States; production in 1957, as a share of total, amounted to 59 percent in Santa Catarina, 37 percent in Rio Grande do Sul, and 4 percent in Parana. Production of clean coal, after washing or handpicking, was 1,466,000 tons in 1957~ A National Coal Plan, approved in 1953 and extended in 1958, aims to establish improved mining methods, increase the production of coking coal, and promote the construction of thermoelectric plants
8 Ibid, P. 5 25
in the mining areas to utilize steam coal.
Petroleum
The Government-controlled company,
Petroleo Brasileiro, S.A. (Petrobras), exercises
a monopoly over the prospecting for or developing
of petroleum deposits. Refineries not in operation
in October 1953 or not authorized before June 30,
1952, also are under Petrobras. The distribution
of petroleum products is handled by private
companies. The production of petroleum on a
substantial scale began only recently and still
is limited to the Bahia Basin. The production
of crude increased from less than 1 million 42- gallon barrels in 1954, to 4 million in 1956,
to 10.1 million in 1957 and to 11.6 million in
January-August 1958. The 1958 production goal
of 17 million barrels is equivalent to 23 percent
of total anticipated petroleum consumption in
the y e ar. Natural gas production in the Bahi a
Basin was 84 million cubic meters in 1956 and
158 million in 1957. Proved reserves of petroleum in the Bahia Basin, a s of Decembe r 31, 1957, were
reported by Petrobras as 418.6 million barrels.
Brazil has a large area of potential sedimentary deposits, and Pe trobras i s exploring and drilling in the middle and lowe r Amazon r egion and in the
Maranhao, Sergipe-Alagoas, Espirito Santo, and
Parana Basins. The production o f petroleum products 26 has risen sharply in recent years, with the coming into operation of new refineries. Refinery throughput increased from 2 million barrels of crude in 1953 to 45 million in 1957. Of the throughput in 1957, 25 million barrels was accounted for by the Petrobras refinery at Cubatao, near Santos, and 2 million by the Mataripe refinery of Petrobras, in Bahia. The remainder of the 1957 throughput was accounted for by privately owned refineries - 10 million by the
Capuaya refinery, near Sao Paulo and 3 million by the Manguinhos refinery near Rio de Janeiro, both of which were inaugurated in late 1954; 2 million by a new refinery, which began operàtions in Manaus in September 1956; 3 million by the
Ipiranga refinery in Rio Grande do Sul;. and 268,000 by the Matarazzo refinery, in Sao Paulo, and the
Uruguaiana refinery, also in Rio Grande do Sul.
Petrobras has under construction in the State of
Rio de Janeiro a refinery with a capacity of 90,000 barrels per day, and the refinery in Bahia is being enlarged to 40,000 barrels capacity. Refinery production in 1957 was: gasoline 17,995,845 barrels; fuel oil 17,277,955 barrels; diesel oil 4,749,456 barrels; kerosene 1,330,019 barrels; solvents
510,605 barrels; lubricating oil 3,898 barrels; liquified petroleum gas 1,914,772 barrels; and 27
asphalt 501,619 barrels. The value of petroleum
production in 1957, estimated by Petrooras in terms
of savings in foreign exchange payments for imports,
was $41 million for crude petroleum and $68 million
for refinery products.
Power
Installed electric generating capacity
of power companies, as of December 31, 1956, was
3,360,000 kilowatts {kw.), of which 2,661,700 kw.,
or 79.2 percent, was hydroelectric power. Installed
capacity, including that of private power plants,
is estimated as 3,586,000 kw. in 1957, compared
with 1,883,000 kw. in 1950. Total electric energy
production is estimated at 15,046 million kilowatt
hours {kw.-hr.) in 1957, compared with 8,565 kw. hr. in 1950.~ Hydroelectric energy is estimated by
the Brazilian National Economie Development Bank
to have supplied approximately 21 percent of the
51.5 billion kw. hr. total of all forms of energy
consumed in 1956, the remainder being accounted
for by petroleum products (49 percent), firewood
{11 percent), coal {9 percent), and other sources
{10 percent). This is nearly double the estimated
27.1 billion kw. hr. total energy consumption in
9 ~, P. 6 1949, of which hydroe1ectric power supp1ied the same proportion, 21 percent, and the percentages from other sources were as fo11ows: petroleum products, 32; firewood, 24; coal, 13; and other sources, 10.
The number of localities served by electricity as of December 31, 1954, totaled lc0 3,669, with a population of 24,053,874. Industrial consumption is estimated to have accounted for
39 percent of the electric energy furnished by the principal producing companies in 1957, residential consumption for 20 percent, commercial consumption for 15 percent, and transportation, public lighting, and other uses for 26 percent.
The hydroelectric potential of Brazil is among the highest of any country in the world.
It is estimated officially as 20 million kw. and unofficially as nearly 30 million. However, the development of power resources from the 1930 1 s until recently lagged so much that the power shortage became perhaps the most severe single handicap to further industrial development.
Generating capacity has nearly doubled since 1950.
Capacity increased by about 75 percent in the years 1954-1957.
The Federal Government and certain
States are active in plans for power expansion.
10 u.s. Dèpartment of Commerce, Basic Data on the Economy of Brazil, P. 12 29
Plans for additional installations to be completed by 1960 would add about 2 million kw., and an additional 3 million kw., for which financing has not been completely arranged, is planned for completion in 1961-65.
Growth of Income and Output
Gross national product increased sharply from 187.6 billion cruzeiros in 1948 to 691.1 billion in 1955 and from 896.1 billion in 1956 to 1~046 billion in 1957. The gross national product for 1958 was estimated to be close to
1,300 billion cruzeirosll at current priees, which would represent an increase of about 23 percent over the 1,046 billion in 1957. Calculated on the basis of 1948 priees, the increase to 1956 was 49 percent.
Net national income at factor cost, in current cruzeiros, rose from 159.5 billion in 1948 to 751 billion in 1956 and an estimated 872 billion in 1957~ Because of the improved terms of trade, the rise in real income was more accelerated than the rise in real product. Total real income increased 61 percent from 1948 to
1956. The percentage share of agriculture and ll Estimate contained in annua1 message to the Brazilian Congress by the President, 1959.
22 Fundacao Getu1io Vargas, Revista Brasi1eira de Economia, December 1957, Rio de Janeiro 1958. 30 livestock rose from 28.1 in 1948 to a high of 29.8 in 1955 but then declined to 26.5 in 1956 and was an estimated 28.1 in 1957. The percentage accounted for by industry increased from 20.9 in 1948 to a high of 24.6 in 1954 and was 22.9 in 1956. (It is of interest to note that, in terms of national product, and excluding Russia and other centrally planned economies, Brazil ranks ninth in the world.)
The development of the national product must be seen against the very high rate of growth of the Brazilian population, which is about 2.5% a year. The Brazilian population, which was forty (40) million in 1939, reached over 60 million by 1958. This representa a population growth of 50% as against a growth of 108% in the real product.
The rapid increase in real output in the period starting with 1948 reflects basically the relatively large scale of public and private investment. According to the best availab1e estimates, investment seems to have been at a level corresponding to about 15% of gross national product during the years 1947 to 1950, going up to 18% in 1951 and 1952 and s1ow1y going down 31
to 15%, 18% and 14% in 1953, 1954 and 1955
respectively. In 1956 the level of 14% was maintained despite great deterioration in the
terms of trade of the country. It may be interesting
to note that, in 1956, 81.2% of total investments in Brazil were made by the private sector of the economy, a proportion which is the highest recorded since 1948.
These statements of the statistics of
growth obviously call for an examination of the
reasons for the growth. Perhaps one of the most
important factors contributing to this growth is the increase in the aize of Brazil's domestic market. The population is now estimated to be
over 63 million. If the rate of growth of the
past decade is maintained for the next eight years,
by 1967 Brazil will have a population of 76 million, and, if the rate of growth of per capita
income is maintained, in the year 1967 the per capita income will be $400. This population will represent one of the biggest national markets in the free world.
This indicates, of course, that Brazil is a rapidly expanding market, one of the few in
Latin America capable of great industrial economies of scale, having established in certain areas the 32 necessary economie and social overhead structure that will enable external economies to be made in the productive processes.
During the early stages of industrialization, the market was largely an urban one, because of the concentration in the cities of the income and wealth created by the traditional export production and its related commercial, financial and governmental activities.
In the northeast of Brazil, it was centred around cities like Fortaleza, Recife, and Salvador, whose wealth depended upon cotton, sugar and cocoa.
In central and southern Brazil it was centred around the governmental and commercial centre of Rio de
Janeiro, the cities of central and southern Minas
Gerais, the city of Sao Paulo, and the trading and export centres of Rio Grande do Sul.
The industrialization process and, above all, the rapid and concentrated urbanization which has gone hand in hand with it, has greatly expanded these markets. At the same time, the spreading of coffee growing to new zones, and the increase in the commercial production of domestic food crops, have been responsible for the springing up and growth of a considerable number of sizable towns in the interior. These are linked to the major 33
industrial centres by a transport system increasingly directed towards integration of the domestic market rather than to overseas trade.
This rapidlygrowing market has now reached sufficient size to permit production, on an optimum scale, of practically all durable and non-durable consumer goods, and quite a few basic producer goods.
However, the growth of real output is by no means solely the result of population growth.
Also of major importance is the rise of a class of industrial entrepeneurs, mainly originating from the group of business men who started operating in the
1930 1 s and who increased the size of their operations and branched out into new undertakings both during and after the war.
In agriculture, there has arisen a new outlook--one which combines the priee consciousness of the large-scale agricultural entrepreneur (long present in export crop production) with much greater alertness towards progressive methods of cultivation.
This alertness has exerted a growing pressure for the establishment of a national fertilizer industry, which is being brought about both as a subsidiary activity of oil refining and as direct production and mining of phosphates. The necessary steps are also being taken for the mechanization of agriculture, not only through the importation of the necessary tractors and other implements, but also through the graduai production of tractors and other 34
implements in Brazil itself.
The recent high rate of output in Brazil
reflects not only high absolute levels of physical
investment but also greater appreciation of the
need for increasing output per man hour and per
cruzeiro invested in plant and equipment.
Particularly since the war there has been
an active interchange of technical information with
foreign countries by means of licensing agreements and
otherwise, under which advanced technical methods have been made available to Brazilian industry.
Progress in the field of training workers and
foremen has also been considerable. It should be
noted, however, that the supply of well-trained
engineers and managers is inadequate when measured
against the needs of the expanding economy.
If any one factor can be said to be most
responsible for the growth of the economy, perhaps
that factor is the stage of inter-dependence which
existe between Brazilian agricultural and industrial
development.
In under-developed economies there exists
a gap between agricultural and industrial development.
The industrial sector is unable to supply the
incentive goods {usually consumer durables) which
would encourage the agriculturalist to greater effort
in order to produce the surplus necessary to buy the
incentive goods. 35
When industry can supply these incentive goods the agricultural sector will increase productivity in order to obtain them. Industry will use agrieultural raw materials as inputs. Agriculture will use equipment manufactured by industry to increase output. They will reciprocally consume each others products and reciprocally supply each others demanda.
Brazil is now beginning to close the incentive gap.
Inflation
No nationwide cost of living data exista in
Brazil. In any case, given the extraordinary disparity within a country of continental aize with various areas at widely different levels of income and general development, such measurement would be of limited significance. The only cost of living index which is reasonably satisfactory from the technical point of view is the one related to wage earners in the city of Sao Paulo. From the base year of 1948
(= 100) this index of the cost of living reached
275 at the end of 1956. This priee increase is of a different order of magnitude from that which was historically experienced by Brazil over a long period, namely, around 2.5% a year compounded. In fact, looking back over the first half of the twentieth century, such priee indicators as exist show that, prior to World War I and from 1925 to 1929, priees 36
were perfectly stable in Brazil. From 1929 to
1935 priees weakened, but between 1935 and 1939 the upswing started, reaching about 30%~which is steeper than the long term trends. The really rapid inflation, however, started only with the outbreak of World War II in 1939.
There is a fundamental difference between inflation in economically advanced countries and inflation in under-developed countries su ch as
Brazil.
In economically advanced countries, inflation is usually associated with high levels of employment of thefactors of production.
Different sectors of the economy try to attract resources away from other occupations by holding out higher money incentives, thus bidding up the priees of those factors and of their products.
In countries such as Brazil, however, inflation does not occur because of high employment but because of imbalance in the development of the economy. For example, horizontal imbalance of industrial development results in bottlenecks and in the artificial bidding up of the priee of alternative resources.
In Brazil, the best example of this imbalance in economie development exists in the 37 field of transportation. The railroad system in
Brazil has not been able to keep pace with agricultural and industrial development, even in the geographie
areas where such development would be both physically
and economically possible. As a consequence of this, there has been a bidding up in the priee of road transportation, and, to a certain extent, in the priee of air transportation.
When alternative and less efficient resources
are used, for lack of a more efficient resource, there resulta a comparatively low amount of productivity
from investment, and thus, high costs and high priees.
When this situation is allied with a
"demonstration effect" as it is in Brazil, where the population is fully aware of the consumption habits of more developed countries, and wher.e the Brazilian population tries to emulate these habits, demand for
consumer goods will also exert inflationary pressure.
If those causes noted above are significant, then the problem is not to reduce investment but to re-direct it into those channels which would eliminate the bottlenecks. This, in effect, is what the government is trying to encourage the market forces to do.
But while the factors discussed above may be significant in the origin of inflation, it must be remembered that industrialization and urbanization have implied the construction of thousands of factory buildings, the acquisition of much equipment and machinery, the movement of hundreds of thousands of people into the expanding industrial zones in the cities for whom housing has to be provided, and the creation of distribution and transportation systems in order to move industrial raw materials to factories.
In addition, external factors which have affected inflation must not be overlooked. From 1939 to 1946 Brazil ran consistent surpluses in the current accounts of its international balance of payments.
The existence of these sizable surpluses, mostly resulting from the sharp reduction of all sorts of imports during the war, are an indication that inflation received a powerful impulse from abroad, at least up to
1947. It is also likely that, during that period, the rise in export and import priees was a major factor in the rise of Brazilian domestic priees.
A last and by no means simple cause of inflation resulting from external factors springs from the fact that the large capital imports made possible by better export priees involve, as a rule, a rather large time lag before their full productivity is felt by the economy as a whole. This time lag is 39
made all the more inflationary because of the need
for counterpart domestic investments.
The situation with respect to the supply
of money and the availability of credit has been
and continues to be serious. Paper money issued in
1958 increased by almost 25 percent from 96.5 billion
cruzeiros to a record 120 billion cruzeiros, as
compared with increases of 16 billion and 11 billion
cruzeiros, in 1957 and 1956, respectively. Total
money in circulation, including short-term demand
deposits, increased by about 20 percent from 291 billion
in 1957 to an amount in excess of 350 billion, as
compared with an increase of 73 billion, or 33 percent, 13 in 1957.
Despite steady and substantial increases
in means of payment, protesta continue against the
lack of credit available, and concern has been
expressed as to where credit facilities would be
found, for example, to finance the sale of the heavy
production of automotive vehicles scheduled for the
next few years.
The closure of severa! banks in Sao Paulo
and Pernambuco, which resulted in runs on other
banks, caused concern in financial circles. Official
intervention resulted in severa! liquidations, and
a large issue of new currency was required to bolster
the banking system.
13 U.S. Department of Commerce, Economie Developments in Brazil - 1958, P. 10 40
The overall deficit of the Federal
Government for 1958 was 34.4 billion cruzeiros,
which was, however, less than the 43.5-billion
cruzeiro deficit in 1957. Revenue totaled 117.8
billion cruzeiros; expenditures totaled 152.2 billion~ The fiscal tiudget for 1959 approved by the Congress allowed for another deficit, of 8.5
billion cruzeiros, to which would be added an
estimated 16-billion to 20-billion cruzeiro cost
for a bonus granted Government workers.
The Minister of Finance submitted a
monetary stabilization program in October 1958
which included provisions for the control of
credit expansion, an increased minimum wage, a
bonus for Government workers, decreased budgetary
deficit, and increased revenues from taxes. The
plan was not enacted by the Congress in its
entirety. An increase of over 50 percent in the
minimum wage was decreed at the end of 1958,
6 months before it would ordinarily have been
due, together with a bonus of 30 percent for
official employees. Revised income, stamp,
and consumption taxes were also enacted. In
addition, emergency priee control regulations were
announced in November in an attempt to limit
priee increases in basic items, especially fooa.
In January 1959 a new program for cutting
Government expenses during the year was announced with a goal of saving 35 billion cruzeiros. 14 Ibid, P. 11 41
Balance of Payments
The Brazilian balance of payments is the victim of structural disequilibrium and rigidity.
Over $300 million are spent yearly for freight and interest on foreign investments, etc.---part of the so-called invisible sector of the current account transactions. The rigidities arise from the fact that
Brazil is committed to spend certain sizable amounts for the importation of essentials without which the economy would break down, (as ~s the case with oil and, to an extent, with wheat and certain basic raw materials) and the amortization of long, medium and short terms loans by international and national public or private institutions.
In 1956, as in 1953, Brazil had positive balances on current account. In 1951, 1952, 1954,
1955 and 1957 it had deficits. But the positive balances have been much more the result of import compression than of increased revenue from exports.
The future balance of payments are going to depend, Qn one hand, on the Brazilian ability to go on compressing imports, and on the other, on a combination of the country's ability to increase exports and attract capital.
But these three components of the future behaviour of the balance of payments are very closely 42 related to each other. The ability to compress imports depends upon the country's ability to produce import substituting goods and resources; the import-substituting production, as well as production for exports, always depends, in sorne degree, on the capacity to import equipment, which, in turn, depends upon foreign investments and credits. But foreign investments and credits flow into a country, only when it is felt that there is reasonable assurance of solvency in that country's foreign trade.
Out of this whole equation looms, of course, the Brazilian ability to export. In terms of the conventional primary exports, such as coffee, cocoa and cotton, the prospects are not generally good. These products, however, should be in a position to guarantee a minimum of around a billion dollars for Brazil annually for the next few years. A lot more could be done and is being done in the field of mineral exports, mostly iron ore, but no dramatic changes are to be expected there.
The possibilities of a break-through in industrial exports for non-dollar markets, mostly Latin American ones, are rather good. Even though it is premature to make quantitative estimates in that sector, this appears to be the biggest hope for expanding exports in the next few years.
On the import~substitution side, the reduction of the freight negative balance is definitely feasible and can easily 43 reach from 50 to lOO million dollars in a short time.
There are here sorne imponderable elements, such as the possible discovery of oil in more dramatic quantities.
This would release almost 300 million dollars yearly for other purposes. The wheat problem should be solved in a few years and will save at least another $100 million in the near future.
With reference to the problem of foreign capital, it is believed that, within the next few years, the economie development program will have aided in the establishment of monetary stability and thus will have provided foreign capital with the assurance it needs. CHAPTER III
THE BRAZILIAN TRANSPORTATION SYSTEM 44
THE BRAZILIAN TRANSPORTATION SYSTEM
CHAPTER III
For a country such as Brazil, a country of great size in which abound natural obstacles to movement, such as rivers and jungles, rapid economie development requires the existence or the establishment of an efficient and economie transportation system. In Brazil, however, the present inadequacy of the transportation system is one of the major bottlenecks of the economy.
There follows in this chapter a brief description of the Brazilian transportation system. The sources of the statistics quoted here are the publications of the United
States Department of Commerce, entitled "Basic Data on the
Economy of Brazil" (Washington, 1959), and "Comparative
Statistics on the Latin American Republics 11 (Washington,
1958). Because of the escarpment which stretches along the Atlantic coast from the northeast to the south, only a few rivera directly connect the interior with the sea.
The railro.ads and roads had their beginnings in the population centres along the coast and originally extended only to the hinterland served by each coastal settlement; for many years, coastwise shipping alone linked the settle ments.
The deterioration of the railroads and the coastal shippingfleet in the years preceding and during
World War II sti mulate d highway cons tructi on a f ter the war and acted as an impetus to the growth of road transportation. The. deficiency of surface transportation
was an important factor als o i n spur~ing the growth of
air transportation in Brazil. 45
Railroad Transportation
Total railway trackage at the end of
1957 was 36,997 kilometres. About 27 percent of the
trackage was located in Sao Paulo and Rio de Janeiro,
23 percent in Minas Gerais, and 10 percent in Rio do
Sul.
Traffic for 1957 was: 347,015,000 passengers (12.2 billion passenger kilometres), 1,305,000 metric
tonnes of baggage (250,551,000 million tonne-kilometres),
39,098,000 tonnes of freight (9.8 billion tonne kilometres),
and 5,041,000 head of livestock (1.9 billion animal kilometres). Tonne kilometres of freight carrred by the
railroads in the years following World War II remained
at a fairly stable level with only a slight annual increase until 1955 when tonne kilometres carried was
9 percent above the figure for the preceding year.
Increases in the rate structure, which stimulated
competition from truck transportation, were to a large degree, responsible for the fact that the increase in tonne kilometres in 1956 was less than 4 percent and in 1957 is estimated as only 1 percent.
The railroads gradually have been nationalized, and the Federal Government in 1955 owned 78 percent of the total railway extension, the States 14.5 percent, and private interests 7.5 percent. The only major private railroad is the Cia. Paulista de Estradas de 46
Ferro, 1,340 miles in length.
The railroads in the east-central and southern
States have been unable to handle the volume of traffic seeking transportation, whereas, in some sections of the country, notably the northeast, the traffic density is extremely light. Re-equipment project loans totalling
$61.1 million were granted by the International Bank for Reconstruction and De~pment and the Export-Import
Bank of Washington. A further Eximbank credit of $100 million, authorized in July 1956 to finance the purchase of United States equipment for rehabilitation of
Brazilian Federal and State railways, is being made available under the supervision of the Rede Ferroviaria
Federal, S.A., an organization designed to unify and improve the administration of the Brazilian Government railways.
Highway Transportation
Highways increased from an estimated 171,900 miles in 1947 to 290,500 miles in 1956. A total of
2,648 miles was paved by 1956 and of the paved mileage,
41 percent was in Sao Paulo and Rio de Janeiro, and
71 percent in these 2 States and Minas Gerais, Parana,
Santa Catarina, and Rio do Sul.
The basic National Road Plan that was developed in 1944 now includes plans foran .additional network of roads totalling 28,684 miles, of which 47
7,372 miles were constructed by the end of 1955. A 5 year construction program for the years 1956-60 under the road plan calls for the construction by the Federal
Government of_ roads totalling 7,800 miles and the paving of 3,500 miles at a total cost of 30 billion cruzeiros.
Good paved highways extend from Rio de Janeiro to Sao Paulo and from Santos to Sao Paulo and inland to beyond Campinas. A highway connecting Rio de Janeiro with
Belo Horizonte was completed in 1957 and a paved Belo
Horizonte-Sao Paulo road is under construction.
Truck and bus transportation is of rapidly increasing importance in Brazil. Because of the inadequacy of rail and water transportation facilities, highway transportation is used extensively for transporting industrial products and raw materials, even bulky items which normally would be carried by rail or sea. An increasing proportion of the traffic consists of heavy trucks.
The number of motor vehicles in operation in
Brazi1 as of December 31, 1957 is estimated as follows:
Buses, 30,701; trucks, 358,496; passenger cars, 395,909; motorcycles, 59,526; and tractors, 43,972.
Water Transportation
Coastal shipping is of paramount importance in providing a transportation link between the south and the north and northeast, and between these regions and the Rio de Janeiro-Santos area. Tonnage transported in the cabotage trade, including some river transportation in the Amazon and Parana Basins, in 1956 totalled
6,525,665 tons compared with 19,669,503 tons in the export-import trade in that year. The cabotage trade consisted of raw materials, 53 percent; foodstuffs,
34 percent; manufactures, 13 percent. Commodities important in the trade include coal, petroleum products, cotton and other fibers, lumber, vegetable oils and waxes, textiles, pig iron and steel mill products, rubber, lime, cement, salt, sugar, rice, coffee, mandioca, wheat and wheat flour, meat, beans, and corn.
The navigable extent of Brazilian rivers is
27,340 miles. The Servico de Navegacao de Amazonia e de Administracao do Porto do Para (SNAPP) is the principal Government-controlled river transportation company. Also Government-controlled is the Servico de Navegacao da Bacia do Prata, which operates in the
Parana River Basin. River service is provided also by a large number of small privately owned companies and individually owned craft.
The total number of ports in use was reported in 1955 as 975 of which 231 were ocean ports and 744 river ports. The Amazon River Basin accounted for 250 of the river ports, and 127 were on rivers of the Parana system.
Only 18 Coastal ports and 1 river port are organized on 49 a commercial basis. In 1956, 47 ports were ports of shipment or destination for goods entering into the foreign trade of the country. Santos and Rio de Janeiro are by far the most important ports, followed in order by Porto Alegre and Rio Grande,
Vitoria (principally iron ore), Niteroi, Recife,
Salvador, Belem and Paranagua.
Although the coastal shipping fleet and river transportation still are inadequate some improvement has recently been made. Among other additions to the fleet, 12 merchant vessels were purchased in 1956 from the u.s. Government, with the assistance of an Eximbank credit, for the use of Cia Nacional de Navegacao Costeira.
Air Transportation
Air transportation has expandèd rapidly, stimulated by the size of the country and the inadequacy of surface transportation. The number of passenger kilometres performed in scheduled revenue domestic and international air services has grown from 1,239.7 million in 1951 to 2,102.8 million in 1957, air cargo freight tonne-kilometres performed have risen from 46.3 million in 1951 to 96.4 million in 1957 and kilometres flown have risen from 76.8 million in 1951 to 121.6 million in 1957. 50
The scheduled airlines offer regular service over a wide area of the country and this is supplemented by special flights by scheduled airlines
and by air taxi service. The Correio Aereo Nacional,
operated by the Air Force, has provided in~tial air
service to remote points in the interior, thus paving
the way for commercial operations.
The following ten Brazilian airlines
presently operate scheduled passenger services, either
domestically, or both domestically and internationally.
CRUZEIRO Servicos Aereos Cruzeiro do Sul
LOIDE Loide Aereo Nacional
Navegacao Aerea Brasileira
PANAIR Panair do Brasil
PARAENSE Paraense Transportes Aereos
REAL Real Aerovias
SADI A Sadia S.A. Transportes Aereos
SAVAG Sociedade Anonima Viacao Aerea Gaucha
VARIG S.A. Empresa de Viacao Aerea Rio Grandense
VASP Viacao Aerea Sao Paulo
The ~ivil airlines are privately owned, with
the exception of VASP, practically all of the capital of which is held by the State and municipal governments of
Sao Paulo. Panair do Brazil is Brazilian-controlled but 51
48 percent of the capital stock is owned by Pan American World Airways.
Airport construction was given a strong impetus by military requirements in World War II.
Many small airports operate in the country and are used by both military and commercial planes. Airports registered up to June 30, 1953, totaled 1,182 of which 78 provided maritime or river facilities. Of the land airports, 1,076 were reported to be ungraded and only
28 were paved. The traffic reported at the 294 commercial airports in 1954 accounted for more than 95 percent of the total passengers and cargo transported in that year.
A program for the improvement of air navigation aids and air traffic control equipment is under way with the aid of United States technical assistance funds, which will supply $1.8 million for equipment and the services of a field staff of United
States technicians.
An analysis of the economie position of the airline industry is presented in the following chapter. CHAPTER IV
AN ANALYSIS OF THE ECONOMIC POSITION OF THE BRAZILIAN AIRLINE INDUSTRY 52
AN ANALYSIS OF THE ECONOMIC POSITION OF THE BRAZILIAN AIRLINE INDUSTRY
CHAPTER IV
Purpose of the analysis
The analysis which follows has been undertaken in order that the position of the Brazilian airline industry may be understood. It is considered essential that a forecast of the performance of the industry be preceded by an examination of the industry itself, since the industry 1 s productive capability for the forecast period will be, in part, determined by its present competitive financial, and technological condition.
Post-war growth of the Brazilian airline industry
The peculiar characteristies of the aeroplane - its ability to fly between points separated by natural obstacles, such as forests, rivers, mountains, etc., and its great speed and flexibility - made it possible for the Brazilian airline industry, in a period of ten years following the second world war, to provide a network of services over nearly all Brazil. The fact that, in
Brazil, there exist great natural obstacles to ground transportation, such as jungle forests, etc. enhanced the advantage which the aeroplane possesses over other transportation media. The great expansion of air services 53 received considerable impetus from the fact that large numbers of transport aircraft became available at the end of the war for sale to industry at low priees, and also from the fact that the cost of surface transportation in
Brazil was high in comparison to surface transportation costs in other countries.
As a result of these favourable circumstances, nearly forty airline companies were formed in Brazil in the years immediately following the war.
The potential advantages offered by the aeroplane as a medium of transportation depends on the high technological standards employed in their manufacture and operation.
The maintenance of an airlin~s servic~therefore, presupposes the existence of sufficient technological resources to maintain high standards of safety and efficiency.
Because of financial and technical deficiencies, the majority of the airlines founded in the post-war years were not in a position to satisfy the high standards required by the nature of their business and by the government agencies responsible for the control of the industry. As a consequence of this, a large number of operators were obliged to go out of business, or to form technical-administrative groups, with the object of over coming their deficiencies through the pooling of resources. 54
Il "Thus, the air transportation industry was modified,_
under the impact of technological and economie
conditions, from a highly competitive, to an aligopolistic
industry. With the introduction of aircraft and
techniques of still more complex operation, thére are
indications that the industry is tending to become a 15 monopoly'!
The changes in the structure of the market,
due to the conditions mentioned above, were further
encouraged by legal measures, put into effect by the
Brazilian Directorate of Civil Aeronautics,_ as soon
as the boom in the industry began to show signs of
levelling off during the late 1940s. By a law enacted
in July,_ 1948, numerous restrictive measures were put
into force concerning the entry of new competitors,_ the
most important of these measures referring to standards
of navigation and maintenance.
The air transportation industry in Brazil today
The air transportation industry in Brazil today
operates as a public utility, thus constituting a type
of legal oligopoly, and, as such, is subject to
governmental regulation. In principle, this regulation
bas, as its main purpose, the development of the national
economy - by encouraging the private operator to offer
increasing services and, at the same time, by s afeguarding
15 National Council of Economies of Brazil, Analysis and Outlook for Brazilian Commercial Air Transportation, P. 5 55
consumers 1 interests through the prevention of
excessive increases in tariff rates. In the specifie
case of air transportation, no company can operate or
cease to operate without the previous and express
authorization of the government. In Brazil, there
are two commissions dealing with this matter: The
Commission of Studies and Concessions of Airlines
(C.E.C.L.A.), which deals with problems referring to
domestic lines, and which functions in conjunction with
the Civil Aeronautics Directorate; and the Commission
of Studies related to International Air Navigation.
(C.E.R.N.A.I.), directly subordinate to the Minister of Aeronautics, dealing with the problems of commercial
aviation in the international sphere.
Other official bodies also have jurisdiction
over certain aspects of commercial aviation in Brazil.
The Civil Aeronautics Directorate deals with questions
of traffic, operations, private flying clubs and legal
and economie matters associated with these aspects
of the industry. This Directorate, having departments
charged with the registration of aircraft, crew licenses, mechanics, radio operators, airport matters and general
administration, is the centralizing or co-ordinating
entity of the air transportation industry in Brazil. 56
The Directorates of Engineering and Air Routes~ which are part of the Ministry of Aeronautics~ are charged respectively wi th the construction of airpo.rts and runways, and with air traffic control facilities and meteannogy. The Commanda of Brazilian air force ~ones
also collaborate in regional problems, such as inspection and maintenance of airports.
The Brazilian government gives direct subsidies to the air transportation industry, which subsidies have, in principle,"the objective of covering small operating deficits of the pioneering lines, which are deemed to
operate in the national interest, and of encouraging a high standard of service in the international lines, so " 16 that Brazil may compete internationally ."
With reference to the domestic 11 p.ioneering"
airlines, a subsidy and a route concession are granted to that company offering the best service, providing that the following conditions are fulfilled:
1. Use of a specifie type of aircraft equipment,
operated within the standards approved by
the Director of Civil Aviation.
2. Standardization of accounts, to permit
efficient audit of the annual financial
statement of the subsidized lines.
16 Ibid, P. 6 57
3. Exploitation of predetermined routes and
segments, except in cases of proven and
accepted inability, i.e. where technical
or other reasons arise which subsequently
render the operation impractical.
In the international field, where the objective of th·~ subsidy is somewhat different, the following are the most ~mportant conditions under w~ch a subsidy is granted:
1. Payment of the subsidy~ which varies with
the type of equipment used, wi.ll be made
on a mileage basis, which is measured from
the last airport of departure in Brazil, and
w~ch will not exceed a distance prescribed
by legislation.
2. The company agrees to use similar equipment
to that of foreign companies exploiting the
same route, to exploit within Brazil and
without subsidy, an annual distance at least equal to the distance subsidized, and to offer a 50% fare discount to members of
the national par~ament and to professional
journaliste. sB
Apart from the direct subsidies mentioned above there exists another type of subsidy, with a special purpose and with a fixed annual value. This is an annual contribution of 450 million cruzeiros to the airline industry for the purpose of enabling airlines to purchase new flight equipment. It was expected that this annual subsidy, although quoted in cruzeiros, would be changed to a hard currency. This would have greatly standardized the importation value of aircraft by eliminating reference to a cruzeiro with a falling international value.
In order to complement the 450 million cruzeiro subsidy, and in an attempt to offset the unfavourable free rate of exchange experienced by
Brazilian airlines when purchasing fuel, aircraft parts and accessories abroad, the airlines have the privilege of using a special exchange rate, by which the U.S. dollar is computed at a favourable rate.
Until October 1958 this rate was 45 cruzeiros to the u.s. dollar when the current free rate was about 150 cruzeiros to the dollar.
In addition, thosé airlines receiving direct subsidies are, as previously noted, considered to be public utility services, and, as such, are exempt from certain taxes. 59
In spite of the fact that the airline industry is today an essential part of Brazil's economy, it appears that the industry is presently in great difficulties.
In 1959, Brazilian domestic airline passenger load factors17averaged between 30% and
40% as against average load factors of between 50% and 60% in Canada and the United States. The economie implications of these low load factors were even more unfortunate than they might seem since competition among airlines had reached such a stage that carriers were giving illegal discounts and advantages to their customers. (One airline was quoted as selling seats for very short domestic routes on a 50% down, remainder-in-60-monthly-installments basis.)
As previously noted, until October 1958, the government supplied the Brazilian airlines with
U.S. dollars- for the purchase of aircraft parts, gasoline, etc. - at the favourable rate of 45 cruzeiros per dollar. In October, 1958, the rate rose to 80 cruzeiros per dollar, and later rose to 100 to one.
These changes were made as part of the austerity pro- gramme introduced by the Finance Minister, and because of them, the airlines find it progressively harder to
17 ntoad-factor" refers to the number of passenger seats sold per aircraft as a percentage of the total number of seats available. 60
meet international competition. The cost of day-to-day operations bas been rising while the government bas been withdrawing the privilege of a subsidised exchange rate.
In addition to the increase in day-to-day operating costs incurred through the decrease in the subsidised value of the cruzeiro.,_ the value of the annual re-equipment subsidy of 450 million cruzeiros has also fallen, because the Brazilian government did not, as was expected, quote this equipment subsidy in a hard currency. Consequently, the value of this annual subsidy has fallen from 10 million United States dollars in 1957 to 4.5 million United States dollars in 1959.
In addition, the general shortage of dollars in Brazil makes it likely that not even this reduced equipment subsidy will become available to the airlines.
As of December 1959, Brazilian airlines had the following jet equipment on order: Panair do Brazil, four Douglas DC-8 1 s; VARIG, two Boeing 707 1 s;and REAL AEROVIAS, four Convair 880 's. As a result of the seriousness of the situation it presently appears that these jet orders may be partly curtailed.
A numbe·.r of suggestions have been made to deal with the criais. First, certain Brazilian government sources,which cannot be accurately identified,have advised that a new company be formed,called Aerobras,which would be a general merger of all Brazilian airlines, domestic 61
and international, in partnership with the government
and undèr its control. An American aviation pUblication
"Aviation Dailyn18- which specializes in air transportation
matters, reported that many Brazilian felt this to be the
worst possible solution because of the poor reputation the government has won for itself as administrator of the
steamship monopoly, the railroads and other public services.
Another suggestion, made by the airlines,
proposed that each airline curtail the number of jets
ordered. It was also proposed that the government select
one type of jet, order three or four of that type, and
then lease them to the airlines.
With regard to the domestic service, Panair
do Brazil (endorsed with sorne exceptions by REAL)
suggested the merger of the airlines into three groups.
Bach group would be assigned a particular region. It
was felt that this would result in a reduction of the
number of flights, and save money, fuel and equipment.
Reports in various aviation publications19
towards the middle of 1959 referred to a confidential
Brazi1ian government report which, it was said, may,
among other things, further serious1y a f fect orders 20 Bra zi1ian c a rriers have place d f or jet aircraft.
18 Aviation Daily, Fehruary 12, 1959
19 Flight Magazine, Ju1y, 1959, and Aviati on Daily, Ju1y 20, 1 959
20 By mid-1960, no change in jet aircraft orders had yet been announced, but the possibi1ity that there would be changes sti11 remained. 62
The survey and report were made by a committee appointed
by the government in the early months of 1959 to take
a close look at the nation's airline industry. Brigadier
Dario Azambuja, Director of the Department of Civil
Aeronautics, headed the committee. Serving with him were
top Brazilian government economie officiais.
Despite the highly secretive nature of the
report, it is believed that the government was told that:
(1) Brazil has too many commercial transports for the
limited passenger market; (2) the excess flight equipment
should be sold abroad; and, (3) the country should be
divided into three zones, with each zone divided percentage
wise among the carriers. For example, the north zone
would be split,. with 70 percent of the traffic going to
Panair do Brazil and Loide Aero,.and 30 percent to REAL,
VASP, CRUZEIRO and VARIG. In the centre zone, 70 percent
of the traffic would be allocated to REAL and VASP, with
30 percent for the others. In the south, VARIG and
CRUZEIRO would be awarded 70 percent and the others would
split up the remaining 30 percent.
Referring to the jet aircraft the carriers have on order, the report recommends that, in order to avoid a public demonstration of financial incapability, the country should get sorne of the long-distance jets on
order- 707's, DC-8's or 880's. However, it .stated that these aircraft should be used only on the routes linking
Brazil with the United States and Europe and that existing 63 aircraft should be used on routes to Argentina and other
South American points.
The report noted that contributing factors to the financial dilemma are rising domestic costs and the falling exchange value of the cruzeiros. There have been steady increases in fares to match the !osses in the value of the cruzeiro, and corresponding demands for new wage~creases.
The report concluded, in short, that all
Brazilian airlines were deep in the red and that a satisfactory solution was not in sight.
Conclusion
The foregoing discussion seems to indicate th~ the immediate outlook for the Brazilian airline industry is rather discouraging.
In a commercial report on the airlines of 21 South America prepared by the writer of this thesis it was suggested that there has been a general over- orde ring of1ong- range aircraft equipment ( e .g, Boeing
707 and Douglas DC~8 aircraft) by airlines in this region.
It appears that the Brazilian government r eport previously r eferred to agrees with this s uggestion.
21 A. E. Rickards , A statis tical analysis of the scheduled airlines registered in the Americas ( ex cluding Cana da and the u.s.A.) and an estimate of their need for aircraft during the period 1959 to 1965 inclusive, (Montreal, 1959) 64
However, any corrective measures taken with regard to these aircraft orders will only affect the international operations of Brazilian airlines, since such aircraft would only be used on overseas routes, and these operations are only a minor part
(in terms of seat-miles flown) of the total of all
Brazilian airline operations.
Therefore, any action which is intended to improve the situation in which the airline industry presently finds itself must, of necessity, be largely directed towards the domestic sector of that industry.
In view of the reported findings of the government-appointed committee and of the general consensus of informed opinion in Brazil, it is believed that action directed at this sector of the industry will, indeed, be taken.
Since the present operations of the domestic sector of the industry are characterised by excessive 2 competitionz and by consequently dubious sales-promotion techniques, it is felt that the main reform needed is the control and reduction of competition, on a selective basis.
This will inevitably mean the elimination of some airlines from some routes, with a consequent reduction in total available capacity offered.
22 For example, until recently, 9 airlines were operating scheduled services on the 227 mile route between Rio de Janeiro and Sao Paulo. 65
However, in my opinion, this reduction in total available capacity should not mean a reduction in total revenue passenger-miles performed. As previously noted, load factors with existing capacity have been very low. With a re-organised and rationalised allocation of routes among airlines, load factors of the reduced total available capacity should be higher, thereby improving the financial position of the airlines.
Yet, even with the increased load factors, it is believed that there will be enough capcity available to absorb growth of traffic in the immediate future.
In spite of the present unhappy condition in which the Brazilian airline industry finds itself, the output -of this industry, in terms of revenue passenger-miles performed, will undoubtedly continue to grow. This situation will exist, it is believed, because of the position which air transportation has achieved in Brazil - a position which makes it a vital element in that countryts economie development. CHAPTER V
DISCUSSION OF METHODS USED IN FORECASTING AIR PASSENGER TRAFFIC 66
DISCUSSION OF METHODS USED IN FORECASTING AIR PASSENGER TRAFFIC
CHAPTER V
Since it is the objective of the thesis to forecast the performance of an industry (the airline passenger transportation industry in Brazil) it is considered desirable to examine briefly the techniques most generally used in industry forecasting before pass!ng on to an examination and discussion of specifie air passenger traffic forecasts.
In his book, 11 Managerial Economies," Joel
Dean 23 notes that there are four basic methods used in the preparation of forecasts. Briefly, they are: surveys of buyers 1 intentions, surveys of opinions of people who know, projections of past patterns, and correlation analyses.
In certain cases, surveys of buyers' intentions and surveys of opinions of people who know may provide a good indication of future demand for an industry's product. However, these cases would seem to be very limited in number, since, in the field of consumers 1 goods, for example, the buying intentions of the population are highly variable, largely because individuals (as consumers) are unable to foresee the choice they will make when faced with the many alternative ways of spending income. Therefore, even should these methods be reliable in certain cases, their very nature
23 Joel Dean, Managerial Egbnomics (New York, Prentice- Hall Inc., 1951) P. 166. 67 suggests that they can only be used in the formulation 24 of short-run forecasts.
Forecasts made by projecting past patterns would seem, in some cases, to be more reliably based than those making use of the two methods previously mentioned. However, this method also can be unsatis- factory, since the projection of a trend through time assumes that the position of the product whose sales are being forecast will remain the same in society, relative to other products, services and income- consuming activities, and it also assumes that other underlying conditions, such as consumers' income, will continue to move in the same manner in the future as in the past. These can be dangerous assumptions, e:specially in a society· that is subject to great technological or cultural change. In certain cases, however, the projection of a trend may provide a . 25 reasonably accurate p1cture of the not-too-distant future. This may be true, for example, for a forecast of the consumption of electricity in a growing economy.
The fourth of the basic methods generally used in forecasting is correlation analysis. This method relates sales of the product to other variables.
(Some of the variables most generally used are population, national income, and labour force.) With this type
24 In the context of forecasting air passenger traffic 11 short-run 11 is defined here as a period not exceeding one year in length. 25 Whether or not a particular forecast is "accurate" depends upon the specifie circumstances. However, 11 accurate" is defined here as being that condition where the achieved figure falls within the range of error allowed for in the formulation of the forecast. 68 of method, the aim is to find a factor, or factors, which
appear to influence sales, and for which data are available.
The factors selected are then studied in order to determine the closeness of thei'r relationship to sales and to each other. Those factors which appear to have an identifiable
and therefore a predictable relationship to sales are selected as the explanatory variables.
The correlation analysis (simple or multiple) may use coincident time series or explanatory variables, in which case, the coincident series must be forecast.
The justification for this analysis must be that the explanatory variables are easier to forecast. Additionally,
as a practical consideration, forecasts of the explanatory variables made by other economists may be available.
Having very briefly discussed industry forecasting methods in general, we shall now go on to examine specifie examples of forecasts of air passenger traffic.
Attempts to forecast the growth of air passenger traffic were first made towards the end of the second World
War. Since that time numerous other forecasts have been made.
There follows here a description and discussion of the methods used in making four representative forecasts.
Since most work on air traffic forecasting has been undertaken in the United States of America, the four forecasts described here are all concerned with United
States domestic traffic. 69
26 The Curtiss-Wright Forecast of 1944
There were obviously inherent difficulties in attempting to forecast air traffic growth in the war and immediate post-war period. However, the attempt was made and it proved to be reasonably accurate considering the difficult circumstances.
Noting that, 11 In the past there has been a very close correlation between total inter-city common carrier passenger-miles and the Gross National Productn,
B. A. McDonald and J. L. Drew assumed that this correlation would continue to exist.
They then went on to assume that the distribution of intercity travel by trip length would be approximately the same in the forecast period as it was in the year
1940, and that intercity air potential travel was made up of trip lengths in excess of lOO miles.
Discussion then centred on those factors in air transportation which could convert some of this potential into actual. Among the factors given prominence was the expected absolute decline in air fares, but, as the report itself notes, 11 With rates at the levels anticipated for the immediate post-war years, however, the securing of a wider general acceptance of air transport will depend
26 B. A. McDonald and J. L. Drew, Air Trans in the Immediate Post-War Period 1 6-1 Buffalo, 1944 70
not on gradua! rate reductions alone but on a combination of lowered rates, continued improvements in service and general breaking down of whatever inhibitions to air travel which may still existrr. 27
The actual forecast of U.S. domestic air passenger traffic growth, however, was not based upon the correlation referred to above, but upon a subjectively adjusted extrapolation of the trend established by past air performance. The correlation between GNP and total inter-city passenger-miles was used, therefore, only to establish the approximate magnitude of the potential inter-city air travel market, and not to forecast the extent to which this potential would be realised.
Table 3 below shows the forecast for traffic up to 1950.
The Curtiss-Wright report notes that, between
1931 and 1940, air passenger mileage growth approximated a straight line on semi-logarithmic paper. That is, the growth rate throughout the ten-year period was almost constant - about 27 percent per year.
27 Ibid, P. 77 71
TABLE 3
Revenue Passenger Miles on Scheduled Airline Flights - 1931 to 1943 Projected to 1950
1 Revenue Passenger-1' :iles ) Thousan ~s :---- Straight Estimated Line Post- Year Actual Trend War
1931 106,442 103,100 1932 127,039 131,100 1933 173,492 166,700 1934 187,859 212,000 1935 279,376 269,400 1936 388,242 342,600 1937 407,296 435,500 1938 476,402 553,600 1939 677,673 703,900 1940 1,041,174 894,800 1941 1,370,000 1,137,000 1942 1,398,000 1,446,000 1943 1,426,596 2) 1,839,000 1944 2,337,000 1945 2,971,000 1946 3,778,000 4,700,000 1947 4,802,000 5,500,000 1948 6,106,000 6,106,000 1949 7,761,000 6,6oo,ooo 1950 9,941,000 7,ooo,ooo
Sourcez Civil Aeronautics Journal, July 1,1941 P.l61.
1) For the years 1931 to 1934 revenue passenger- miles are not reported separate1y from total passenger-miles inc1uding non-revenue. 2) Estimated from first eight months actua1. 72
It will be noted that for 1946 and 1947 the
forecast ( 11 estimated post-war 11 ) gives higher figures
than a projection of the past trend would have justified.
This is considered reasonable by the authors of the
forecast, bearing in mind the peculiarities of the time
at which the forecast was m.ade (1944), when there was
expected to be a large number of war-surplus aircraft
available, and an extension of the war-time enthusiasm
for aviation into actual peace-time air travel.
The forecasters believed that the rate of
growth in the years 1948, 1949 and 1950 would fall
considerably below what they considered would be the
11 artificial 11 growth rate levels to be experienced in
the immediate post-war years as a result of the war
time stimulus. In fact, however, as will be seen in
that section of this chapter which compares forecasts
with actual traffic achieved, the actual figures for
U.S. domestic air passenger travel in 1950 were almost
half-way between the figure shown in the column headed
"estimated post-war 11 and the figures shown in the
column headed 11 straight line trend".
In summarising, the method used in this
fore c ast can be said to be an adjus t e d projection o f
the trend established during the pre-war and war years into the forecast period. Whilst the authors of this 73
forecast noted that a close correlation existed between
inter-city common carrier passenger miles and GNP,
they merely used this correlation in order to estimate
the magnitude of the potential inter-city air passenger
market, and not to forecast the extent to which this
potential would be realised.
The authors of this forecast implied that,
in the 1946-1947 period, it would be the growth of
available capacity (supply) that would determine the
level of passenger miles to be performed. The market
would be a 11 seller's11 market, . due to the particular
circumstances that the authors expected to exist in
the immediate post-war period. As noted above, the
forecasters did not believe that a high rate of growth
would be maintained beyond 1947 and their forecast for
the years 1949 and 1950 was, therefore, below the straight
line trend.
Thus, although the forecast method used can
be described basically as an extrapolation of the trend
established by past performance, the forecasters did not expect actual traffic to adhere strictly to the trend
line. Variations in the trend line were introduced by
the forecasters to reflect their expectations of the market conditions. 74
Criticism of the method used in making this forecast can be directed at its authors on the grounds that, having noted the existence of a correlation between inter-city passenger miles performed and GNP, and having subsequently determined the magnitude of a somewhat arbitrary air passenger
11 potential", they then pi"oceeded to make their forecast by subjectively modifying the trend projected from the past into the forecast period.
Ho~ever, it is felt that the conditions pre vailing at the time the forecast was made (1944) rendered forecasting of any kind a hazardous occupation, since there was no reasonably reliable indication as to theoonditions which would prevail in the post-war
}"orld. Thus, I helieve that the authors of this forecast may be excused th~ir lack of complete objectivity.
In fact,it is difficult to see how one could have been more objective~in view of the uncertainties the economy faced entering the post-war period.
The Curtiss-Wright forecast dealt with both
U.S. domestic and international traffic. However, both forecasts were similar in nature, and it i s thought that discussion of either one will serve to illustrate and describe the method used. 75
28 The Port of New York Authority Forecast of 1950
This Port of New York Authority (PNYA) forecast considered several factors in arriving at final estimates, the most important of which were: national income, traffic trend, {i.e. the distribution of the total inter-city passenger traffic among the various transportation media), technological-service
(e.g. dependability, safet~ frequency of service, speed, etc.) and traffic pattern {i.e. length of trips).
The report commences with the statement that "The transportation future depends basically on what the American economy will be like in the years 29 ahead 11 •
This statement is enlarged by noting that past trends indicate that traffic volumes and economie activity are closely related. The output of goods and services and the activities of those who produce and consume them are predicated on the ability of trans- portation facilities to overcome the obstacles of time and distance.
28 G. H. Aldrich, A. L. Hart, J. D. McGoldrick, W. Owen, L. C. Sorrell, Air Traffic Forecast - 1950-1980, {New York, 1950).
29 Ibid, P. 1 76
One of the first tasks that the PNYA forecast, therefore, considers necessary for forming
an estimate of tomorrow 1 s traffic is to arrive at some concept of the level of national income in the years ahead; that is, to arrive at some concept of the volume of goods and services that will be produced, and also of the number of people who will be demanding such goods and services.
Table 4 below (reproduced from the report) shows the population and national income statistics for the period 1929 to 1949, with a projection through to 1980. The population projection appears to be a straight line on an ordinary (non-logarithmic) grid, thus indicating a falling rate of population increase, whilst the national income projection is neither a straight line projection on an ordinary grid nor a s traight line projecti on on a s emi-logari thmic grid, but rather a projection which shows an alternately rising and falling rate of income increase. 77 TABLE 4 TRENDS IN POPULATION AND NATIONAL INCOME
1929-1980
N a t i o n a 1 I n c o m e 194.9 Dollars 1931\-1939 Dollars Populatio) Total Per Total Per Year (Millions (Billions) Capita (Billions) Capita 1929 122 $121 $ 990 $ 71 $ sss 1930 123 106 863 63 511 1931 124 91 738 54 436 1932 125 72 579 43 342 1933 126 73 577 43 342 1934 126 86 680 51 402 1935 127 98 769 ss 455 1936 128 110 862 65 510 1937 129 121 941 72 557 1938 130 113 871 67 515 1939 131 123 912 73 557 1940 132 137 1,039 81 614 1941 133 167 1,253 99 741 1942 135 198 1,471 117 870 1943 137 230 1,687 136 998 1944 138 246 1,779 145 1,052 1945 140 239 1,711 142 1,014 1946 141 218 1,511 129 912 1947 144 215 1,191 127 883 1948 147 222 1,512 131 894 1949 150 222 1,477 130 867 1950 152 223 1,467 132 868 1955 158 241 1,524 142 901 1960 164 265 1,616 157 956 1965 170 293 1,725 173 1,020 1970 176 331 1,882 196 1,113 1975 181 368 2,034 218 1,203 1980 186 416 2,239 246 1,324 78
It is to be noted that the population forecast for 1960 shown above is considerably lower than the population estimate for 1959 contained in the 1959
"Statistical Abstract of the United States", published by the u.s. Department of Commerce. According to this Department of Commerce estimate, the population of the u.s. reached 176 million by March 1959, 30 twelve million more than that forecast by the Port of New
York Authority for the year 1960. Similarly, the
PNYAts forecast of National Income has so far proved to be too conservative. According to the above- mentioned Statistical Abstract, National Income in
1958 was estimated to be $360.8 billion current 31 dollars • Deflating this figure to the same 1949 dollars used in the PNYA forecast results in a 1958 actual National Income of $299.8 billion, which is some $35 billion higher than the PNYA forecast of
265 billion 1949 dollars for 1960.
Having presented these statistics on population, national income and per capita income, the PNYA report goes on to discuss the relationship between per capita income and per capita passenger mile travel, stating that total inter-city travel tends to increase at a slower rate than national income but
30 - u.s. Department of Commerce, Statistical Abstract of the United States - 1959, p. 5. 31 - Ibid, p.305 79
32 faster than population. Table 5 below (reproduced
from the PNYA report) shows the relationship between
per capita income and per capita travel. For example,
between 1933 and 1940 per capita income increased 80%,
compared with a passenger-mile increase of only 48%
(and a population increase of 5%). Stating that the
data for the period 1929 to 1949, exclusive of the
war years, illustrated the existence of a simple
correlation between per capita travel and national
income, the authors of the PNYA report extrapolated
the volume of per capita travel to forecast the u.s. domestic inter-city passenger-miles through to 1980.
Table 5
NATIONAL INCOME AND TOTAL INTER-CITY PASSENGER TRAVEL 1953-1980
PER CAPITA PER CAPITA INTERCITY NATIONAL INCOME PASSENGER TRAVEL Per Cent Per Cent Total Intercity 1935-1939 Increase Passenger Increa se PASSENGER Miles Year Dollars Over 1933 Miles Over 1933 (Millions) 1933 $ 341.6 % 610.5 % 76,680 1937 556.7 63.0 840.5 37.7 108,261 1940 614.4 79.9 906.1 48.4 119,605 1949 866.7 153.7 1,130.9 85.2 169,642 1950 867.8 154.0 1,100.0 80.2 167,200 1955 901.3 163.8 1,150.0 88.4 181,700 1960 956.1 179.9 1, 200.0 96.6 196,800 1965 1,,20.0 198.6 1, 250.0 104.8 212,500 1970 1,113.1 225.8 1,275.0 108.8 224,400 1975 1,202.8 252.1 1,325.0 117.0 239,825 1980 1,324.2 287.6 1,350.0 121.1 251,100
32 - Compa re this s tatement wi th that made on the s ame subject in the Civil Aeronautics Administration forecas t, discussed on p a ge 9 2 of this the sis. 80
Having arrived at this forecast of total
inter-city passenger-miles to be performed, the
authors of the report then turned to an examination
of the distribution of these passenger-miles among
the various transportation media. Noting that
private automobiles accounted for 66% of total
inter-city passenger-miles in 1949, the report
concludes that the existing rate of growth in inter-
city automobile travel will persist and that,
therefore, between 1949 and 1980, automobile travel
will show a better than 65% increase, but that there will only be a 10% increase for the common
carrier group in this same period. The forecast
is given in Table 6 below:
TABLE 6
ESTIH~TE OF TOTAL INTERCITY TRAVEL Millions of Passenger Miles 1950-1980
Year Total Common Carrier Automobile
1949 169,642 57,600 112,042 1950 167,200 53,960 113,240 1955 181,700 53,720 127, 980 1960 196,800 54,940 141,860 1965 212,500 57,120 155,380 1970 224,400 59,312 165,088 1975 239,825 61,178 178, 647 1980 251,100 63,054 188,046 Per Cent I ncre ase 1980 over 1949 48.0% 9.5% 67.8% 81
In the words of the report itself, "These estimates reflect the dominant position attained by the private automobile during the late twenties and thirties. Over the course of the past several decades, the common carriers have been relegated to a secondary role. By the late thirties almost 75 per cent of all inter-city travel was by private automobile. In 1949 automobile travel accounted for 66 per cent of total inter-city passenger- miles, representing a substantial, but by no means complete recovery from wartime levels when the automobile carried only 32 per cent."33
However, like the forecasts of population and national income, the forecast for common carrier passenger-miles also appears to be too conservative. Whereas the forecast for the 1960 passenger-miles to be performed by common carriers
(railroad, bus and airlines) was put at 55 billion, in actual fact this figure was already over 79 34 million in 1957, thus showing that the residual
10% increase forecast for the common carrier group by 1980 was a serious under-estimate, as was the absolute figure for total passenger miles.
33 P.N,Y.A. Forecast, p.8
34 U.S. Department of Commerce,op.cit.,~565 82
The next approach considered by the authors of
this forecast is the technological-service approach.
Under this heading there is discussed dependability
of operation, safety, frequency of service, persona!
comfort, competitive fares, and speed. The comments
made are, necessarily, of a very general nature,
noting advances and expected advances in aircraft
design and operating techniques which will have an
effect on operating costs ( and presumably, therefore,
on fare levels ) and aircraft performance. The
authors conclude that the development of aviation
will enable air transportation to strengthen its
competitive position in relation to surface
transportation.
The report then proceeds to an examination of
the most important aspect of the traffic pattern
from the point of view of domestic air transportation.
This factor, it concludes, is the length of the trip.
As the report states:
"In the present inquiry, length of haul - whether the traffic be passenger, cargo, or mail - has assumed a most important position because of the advantage air transportation possesses in serving long haul, as contrasted with its equally clear disadvantage in serving short haul traffic. This circumstance necessarily influences the division of traffic among the various modes of transportation as well as the development of new traffic brought into being through the considerable time saving made ~ossible by air transport.n35
35 - P.N.Y.A. Forecast, p. 62 83
The PNYA report notes that, in contrast to the assumption made in the Curtiss-Wright forecast, the mileage pattern for inter-city travel reveals that there was an increase in the average length of trip performed by rail between 1940 and 1947.
As rail transportation is considered to be the greatest domestic commercial competitor of air transportation, the increase in average rail trip length is regarded as significant, since it indicated both a shift of short-haul traffic from rail to the bus and the automobile, and also the development of an increasingly large market for long haul common carrier travel. It is from this developing market that the report expected air transportation to make its gains, penetrating the market previously held by rail transportation.
The authors of the report then note that about 35 per cent of domestic rail/air travel is on stage distances in excess of 1000 miles, and the statement is then made that, because of the trend towards an increase in the average trip length, by
1980, nearly 50% of inter-city passenger-miles performed would be on stage distances in excess of 1000 miles. Because of this increase in average 84
trip length and because of air transportationts
inherent advantages, the report expressed the
opinion that virtually all of these passenger-
miles would be performed by air. The authors of
the report further opined that, on stage
distances between 500 and 1000 miles, air would
perform 75% of the passenger-miles, and on stage
distance between 150 and 500 miles air would
perform 50%.
Using the forecast of total common
carrier inter-city passenger-miles contained in
Table 6 on page 80, the authors of this forecast,
bearing in mind the qualitative aspects of each
medium of transportation, then made a purely
subjective calculation of the shares of this
forecast total which would accrue to each of the
common carrier media during the forecast period.
The resulting forecast is shown in Table 7 below:
Table 7 ESTIMATE OF INTERCITY COMMON CARRIER TRAVEL (Millions of Passenger Miles) 1950-1980 Year C8~bn Carrier Rail_. Air Bus 1949 57,600 29,500 6,600 21,500 1950 53,960 25,906 7,300 20,754 1955 53,720 22,904 11,066 19,750 1960 54,940 21,092 14,332 19,516 1965 57,120 19,843 16,707 20,570 1970 59,312 18,331 19,333 21,648 1975 61,178 17,053 21,500 22 , 625 1980 63,054 17,000 22,353 23,701 Per cent change 1980 over 1949 +9.5% -42.4% +238. 7% +10.2% 85
To summarise the method used in this PNYA forecast, both correlation analysis and subjective reasoning were used. Forecasts of national income and population were first obtained and then per capita national income was related to per capita inter-city passenger miles. Using this relationship between per capita national income and per capita inter-city passenger-miles, a forecast was made of the latter. By multiplying this per capita inter~city passenger-mile figure by the forecast population figure, a forecast of total inter-city passenger-miles was derived. At this stage, a rather subjective division of these inter city passenger-miles between automobiles and the total of the common carriers was made. There followed general discussion of aircraft characteristics vis-à-vis those of surface transportation, and discussion of the development of greater average trip lengths. As a result of points raised in these discussions the authors of the forecast made a purely subjective division, between rail, air and bus, of the figure forecast as the total passenger-miles to be performed by all common carriers. 86
Since this forecast, in its initial stage, is based fundamentally upon forecasts of national income and population which have subsequently proven rather inaccurate, it is not surprising that the end result the forecast of air passenger-miles to be performed is also inaccurate. (Comparison of forecasts with results achieved will be found on page 108of this thesis.)
It is not apparent from the text of the forecast itself whether the national income and population forecasts were made by the authors of the report themselves, or whether these forecasts were obtained from another source. It is considered most probable, however, that the forecasts were obtained from an outside source, and, if this is the case, criticism of these forecasts, on the grounds that they subsequently proved to be inaccurate, cannot legitimately be levelled at the authors of the PNYA forecast.
However, legitimate criticism can be directed at those parts of the PNYA forecast where subjective divisions of total inter-city passenger-miles were made. It does not seem logical, des irable or necessary that the end 87 result of this PNYA forecast should have been arrived at by taking one long leap in the dark, which was essentially what the forecasters did when they made the subjective jump from the estimated total of inter-city passenger-miles to be performed by all common carriers to the individual totals for each of the common carriers.
Such a process did not logically follow upon the processes which had preceded i~ and it abandoned any pretence of objectivity. Such a process was not necessary, in my opinion, since the authors of this f orecast could have complemented the forecast of total inter-city common carrier passenger-miles with individual forecas ts for each of the common carrier media.
These forecasts could, for example, have been based upon a possible correlation between each individual medium and per capita income. The sum of thes e individually constr ucted forecasts could have been compared to the total forecast, and adjustments made if the results were not reasonably close.
The method just described would have eliminated the highly s ubjective feature of the
PNYA forecast and made its f i ndings more defensible . 88
The U.S. Civil Aeronautics Administration Forecast of 195536
This report, which was prepared as a staff study for official use in planning, contains a forecast of u.s. domestic airline traffic.
The authors of this report, before forecasting the volume of domestic airline traffic, present a forecast of total inter~city passenger traffic, based upon considerations of the growth in gross national product.
Table 8 below, reproduced from the Civil
Aeronautics Administration (CAA) forecast,shows the growth of GNP during the period 1929 to 1954 with a projection through to 1965.
36 - U.S. Civil Aeronautics Administration, Civil Air Traffic Forecasts 1960~1965, (Washington, 1955) 89
TABLE 8 GROSS NATIONAL PRODUCT 1929 - 1965
Gross National Produet Year Current 1954 Priees 1 priees {billions of dollars) 1929 104.4 181.0 1933 56.0 125.8 1939 91.1 189.4 1940 100.6 205.9 1941 125.8 236.7 1942 159.1 264.7 1943 192.5 294.3 1944 211.4 317.1 1945 213.6 311.8 1946 209.2 279.3 1947 ~32.2 278.3 1948 257.3 292.4 1949 257.3 291.4 1950 285.1 318.5 1951 328.1 340.2 1952 346.1 353.2 1953 364.9 368.5 19541 357.1 357.1 1960 - 436.0 1965 - 535.0
SOURCE: Economie Report of the President; January 1955 1. Preliminary 90
The forecast of 1960 GNP at 436 billion
1954 dollars appears to be reasonable, since
the 1959 "Statistical Abstract of the United
States" shows that the 1957 GNP in 1954
dollars was 407 billion and was growing at a
rate consistent with the CAA forecast.
The next step in the CAA forecast is an examination of the relationship of GNP to expenditure on domestic travel. The report notes that historically: "the volume and amount spent for transportation services has been closely correlated with our level of economie activity.n37
37 - Ibid, p.5 91
Table 9 below (reproduced from the report) indicates that,between 1946 and 1954,domestic inter- city travel expenditures approximated 5 per cent to 6 per cent of total gross national product, with the proportion spent on travel increasing at a slightly faster rate than GNP.
TABLE 9
RELATION OF DOMESTIC TRAVEL EXPENDITURES TO GROSS NATIONAL PRODUCT, 1946 1954
Gross National Domestic Total Year Product- Travel Percent intercity Current Doliars Expenditures .of passenger- (billions} (billions) GNP miles (billions)
1946 $209.2 $10.7 5.1 344.6 1947 232.2 12.4 5.3 342.9 1948 257.3 13.9 5.4 352.2 1949 257.3 15.0 5.8 375.5 1950 285.1 16.0 5.6 393.3 1951 328.2 19.3 5.9 441.8 1952 3j6.1 20.8 6.0 465.0 1953 364.9 22.2 6.1 479.3 1954 357.1 22.7 6.4 501.7
SOURCE: Economie Re~ort of the President, January 19551 CAB and ICC Statistics 92
Thus, noting that growth in total inter-city travel is related to economie growth, the report continues: "since 1946, GNP in constant priees has risen nearly 30 percent. Total inter-city travel has grown more rapidly - from 344.6 billion passenger-miles in 1946 to 501.7 billion passenger- miles in 1954, a gain of 46 percent. During the same period, per capita inter-city travel advanced
27 percent and inter~city passenger-miles for each
$1,000 of GNP increased nearly 14 percent. On the basis of these historical trends and relationships and, to be conservative, assuming that inter-city passenger volume will increase only in proportion to increases in the GNP, total inter-city travel may be expected to reach approximately 750 billion passenger~mile s by 1965 (a 50 percent increase over 38 1954)". In order to determine how much of this 750 billion represents air potential, an analysis is then made of the make•up of inter-city travel. How is it distributed among the various media of transportation ?
The CAA included Table 10 below in their report in order to s how this dis tributi on.
38 Ibid, p.6 TABLE 10 DOMESTIC INTERCITY PASSENGER-MILES BY MODE OF TRAVEL AND CLASS OF SERVICE CALENDARYBARS 1946-1954 (In millions)
Mode and Class 1946 1947 1948 1949 1950 1951 1952 1953 1954 Total 344,613 342,895 352,192 375,512 393,328 441,873 465,004 479,323 '501,677 Total Common Carrier 91,043 69,937 64,779 59,738 55,989 62,549 62,946 61,323 58,597
Scheduled Air Carrier lJ ls210 6.06J 5:~210 6.zo5 1o.soo 12.461 14,688 16 1 626 Regular service 5,910 6,063 5,905 6,454 9,226 10,962 10,962 11,375 Coach service ------5 251 ~1,056 1,274 3, 726 3,726 5,321 Class I Line-Haul RailwaysJ/ 58:~840J2a226 JS:aJJO 22,622 26,.z81 22aZ.lO 22 1 262 26:~20524:a§32 First-class service 19,801 12,261 11,015 9,349 8,338 10,226 9,504 7,950 6, 50 Coach service 39,039 27,665 24,315 20,273 17,443 19,524 19,758 18,955 17,689
MOtor Carriers11 26 2 223 2J:a248 2Ja522 22,4;11 21 1 254 22 1 222 21:~223 12,.ZJO lZ 1 J 62.4.1
Private Automobile!/ 25_3_,52_0 2Z2:~25828Z 1 42J Jl6,ZZ4 402 2 058 418,000 44Js080 JZ2a324 1 :uz,JJ2 \C (..; Percent air to total 1.7 1.8 1.7 1.8 2.0 2.4 2.7 3.1 3.3 Percent air to total conunon carrier 6.5 8.7 9.1 11.2 14.2 16.8 19.8 24.0 28.5 Percent air to total rml 10.0 15.2 16.7 22.6 29.7 35.3 42.6 54.6 68.0 Percent air to first- class rail 29.8 49-4 53.7 71.7 85.2 102.7 131.1 184.8 243.7 ~- 11 Scheduled operations of domestic trunk line and local service carriers (irregular carrier data not available). y Includes Pullman Company and excludes commutation. Excludes local, suburban and charter service. ~ Estimated intercity passenger-miles.
Source: Civil Aeronautics Board, ~~line Statistics Handbook and C.A.B. and I.c.c. statistics. 94
In commenting upon this table, the authors of the report state "Indications are that the pattern of the past few years will not change appreciably," following this with the statement that "Total common carrier traffic is expected to show a moderate rise to approximately 66 billion passenger-miles through
1965 largely as a result of further growth in air travel." 66 billion is only 8.8 percent of the 750 billion passenger-miles that the CAA forecast as the total of ali inter-city passenger-miles for 1965.
In 1954 the common carriers share of the total inter city travel market was 12 percent. It is apparent, therefore, that the "pattern of the past few years" referred to above is the increasing share of the total inter-city travel market which has accrued to the automobile and the correspondingly decreasing share which has accrued to the common carriers.
In 1954 air transport's share of the common carrier total was 29 percent. The CAA forecast that this would rise to more than 50 percent by 1965. The reasons for this forecast were as follows:
"An expansion of this magnitude assumes
that air transportation will not only
increase its long-haul traffic but will 95 also make a substantial penetration of the medium and short-haul travel markets. The airlines already dominate the long-haul market - in
1953 nearly 54 percent of common carrier passengers making trips over
500 miles moved by air. Moreover, long-haul traffic has become an increasingly important segment of airline business - the proportion of airline passengers making trips of 500 miles or longer has increased from 25 percent in 1941, to 35 percent in 1946 and again to
43 percent in 1953. Although long-haul operations should continue to expand, because it is in this market that the speed advantage of air transportation is most apparent, it is expected that a large part of the estimated increase in airline traffic will be in me dium and short-haul operations.
These markets represent the great 96
bulk of common carrier travel and
so far the airlines have captured
only a minor portion of this
traffic. In 1953, approximately
89 percent of ali common carrier
passengers made trips of 250
miles or less while 96 percent
traveled distances of 500 miles
or less. The airlinest share of
this traffic bas been small - less
than one percent of ali common
carrier passengers travelling~up
to lOO miles, 9 percent of those
travelling 101 to 250 miles, and
25 percent of those travelling 251 to
500 miles.
Increased airline penetration of
the medium and short-haul markets
may largely be expected to take
two forms: (1) expansion of
coach services to these mileage
brackets, and (2) inauguration
of inter-city helicopter service."39
39 Ibid, p.8 97
Stating that inter-city passenger travel has
increased at a rate faster than the increase in GNP
and population, the report notes that air travel has
become an increasing portion of total inter-city
traffic and that there is no reason to believe that
this trend will change. "Thus", as the report says,
"there is no lack of room for the expansion of air
traffic."
There follows a few explanatory notes, and
then the forecast, reproduced in Table 11 below, is
given. TABLE 1.1
DOMESTIC AIRLINE PASSENGER TRAFFIC AND CHANGE FROM PREVIOUS YEAR
1 Revenue Average Revenue pas sengers passenger-miles trip (millionsl (billions) length Year Number % Chanee Number % Chane:e (miles) 1946 12.2 - 5.9 - 487 1947 12.9 5.7 6.1 3~4 474 1948 13.2 2.3 6.0 - 1.6 454 1949 15.1 14.4 6.8 13.3 448 1950 17.3 14.6 8.0 17.6 461 1 951 22.7 31.2 10.6 32.5 466 1952 25.0 10.1 12.5 17.9 501 1953 28.7 14.8 14.8 18.4 514 1954 32.3 12.5 16.8 13.5 518
1960 55.0 29.0 520 1965 Low 68.0 35.5 520 Inter 70.0 37.0 520 High 72.0 38.5 520
SOURCE : 1946-1954 da ta f rom Office of Carrier Accounts and Statis tics , CAB. 1960 a nd 1965 e s tima t e d by CAA Office of Planning, Research and Development. 98
The 1960 forecast of revenue passengers is a simple projection based on the average rate of growth between 1950 and 1954 of 3.8 million passengers each year. The 1965 forecast is based upon an annual increase of 3 million passengers between 1961 and
1965.
Passenger-mile forecasts of 29 millions for
1960 and 37 billion for 1965 are based on the 1960 and
1965 forecasts of 55 million and 70 million passengers for these years respectively and on an average trip length of 520 miles.
The method used in this CAA forecast can be summarised as follows:
a) A relationship between GNP and
domestic trave1 expenditure
(and, therefore, to inter~city
passenger-miles) is noted;
b) it is assumed that inter-city
passenger volume will increase
on1y in proportion to increases
in GNP (a forecast of GNP for
the years 1960 and 1965 is
inc1uded in the report);
c) a forecast of the inter~city
passenger-miles to be performed
by 1965 is, therefore, made. 99 d) the distribution of inter-city
passenger travel among the
various transportation media is
examined; e) the statement is made that the
common carriers' share of the
total inter-city passenger
miles to be performed in 1965
will be approximately 66 billion. f) it is further stated that, by
1965, more than 50 percent of
the common carrier total will
be performed by air, this
expectation being based not only
upon the natural growth of air
travel, but also upon the ability
of air to penetrate both the
surface carriers' long-haul and
its short-haul markets; g) the growth of domestic airline
passenger traffic, in terms of
passengers carried, is
projected through to 1965; h) by applying average trip lengths
to the forecast figures for
passengers to be carried, the
forecast of the number of
passenger-miles to be performed
is obtained. lOO
It is considered that there are several
points about the methods used in this CAA forecast
which deserve comment.
There appears to be no reasoned
explanation for the forecast of total common
carrier inter-city passenger miles to be performed
in 1965. There is just the statement of the
expectation that 66 billion passenger-miles will be
performed in that year, explained only by the remark
that the rise represented by this figure over the
1954 figure is "largely as a result of further growth
in air travel." It is felt that this is an insufficient
explanation of the forecast.
With regard to the figure of 66 billion
itself, as previously noted, it suggests that an even
smaller percentage of the total inter-city tra~el
market will be available to the common carriers in 1965
than was available in 1954. In 1954, 12 percent of
all inter-city passenger-miles performed were
performed by common carriers. According to the CAA
forecast, this figure will drop to 8.8 percent in
1965, leaving 91.2 percent to be performed by the private automobile. On the face of it this appears
somewhat startling, but, according to statistics
contained in the 1959 edition of the previously mentioned "Statistical Abstract of the United
States," between 1954 and 1957 (the latest year for which figures are available) the percent of inter- city passenger-miles performed by private automobiles 101
rose from 87.79 to 88.67, and the rise during this period was steady and uninterrupted.
Following the statement regarding the
66 billion common carrier passenger-miles, there
is a statement that over fifty percent of these
passenger-miles will be performed by air, as against
the twenty-nine percent of the common carrier total
performed by air in 1954. In support of this
assertion certain figures showing the percentages of
common carrier passengers making trips of various
lengths are given, in order to indicate the magnitude
of the medium and short-haul market potential which
will become available to the air carrier. However,
it appears that the figure of fifty percent is a
purely subjective estimate, apparently not based
upon any objective calculation.
When the actual forecast itself is examined
it can be seen that the method used to formulate it is
one of a simple projection of the past trend. It appears, therefore, that all the calculations, discussions and subjective divisions which p-ee:ceded it were undertaken merely in order to provide justification for this simple projection. 102
Unlike the method used in the Curtiss~
Wright forecast, where a correlation between GNP and inter-city common carrier passenger-miles was used to provide the framework in which it was possible to forecast a specified growth in air travel, this CAA forecast appears to have used the aforementioned correlation (and a great deal of subjective reasoning) in order to arrive at a forecast figure of roughly the same magnitude as that arrived at by a simple projection of the past trend.
To illustrate this, it is noted that the authors of this forecast, following upon their calculation of the total inter-city passenger-miles to be performed by all media, based upon the GNP inter-city travel correlation, subjectively estimated that approximately 66 billion inter-city common carrier passenger-miles would be performed in 1965, and they further subjectively estimated that "more than 50 percent" of this total would be performed by air. Their intermediate trend projection forecast for 1965 is 37 billion passenger-miles, which is 4 billion "more than 50 percent" of 66 billion. 103
The way in which the results of the three
methods used in this forecast {correlation analysis,
subjective reasoning, and projection of past trends)
appear to support each other leaves one with the
suspicion that perhaps the subjective calculations
were made with an eye on the projected trend
forecast. 40 The Canadair Forecast of 1956
K. H. Larsson of Canadair, being
dissatified with the methods used in previous
forecasts, decided against the method of relating
the growth of transportation as a whole to such
economie indicators as GNP, population, labour force,
etc., and then deciding what portion of the total
trans portation pie would be taken by each of the
transportation media.
Instead, the method used in this forecast
was an analysis of the past growth history of each medium individually. This analysis took into consideration the factors in transportation which affect its marketability - speed, reliability, convenience, safety, cost.
The f orecast of each medium's growth was then made by taking into consideration the past trend and any changes in the above noted factors which were expe cted to take place.
40 K. H. Larsson, A Critical Review of Earlier Forecas ts of Air Tra ffic a nd a New Approach,(Montreal, 1956). 104
Discussion of the Canadair method will be
confined here to the air traffic forecasting method,
although such discussion would apply equally well to
the forecasts for the other media.
In the field of speed, it was noted that the
aeroplane has a great advantage over all other forms
of transportation and that this advantage would
increase in the future. Reliability was discussed,
noting that, in air transportation, there still is
great room for improvement, but also noting that this
improvement is being made.
Under the heading of convenience, it was
remarked that air travel generally affords greater
comfort than other media, and that frequency of
operation is increasing greatly, thus offering the
potential traveller service competitive with other media.
The safety record of the airlines has
improved greatly over the past years and the report
notes that, in terms of passenger fatalities per
100 million passenger-miles, air transport
s tatis tics are approaching the s afety level of
surface transport. 105
Under the heading of cost, the report
states that there is no technological improvement in
sight which could lower the cost of rail or motorbus
transportation. This is not true of air transportation, however. u.s. domestic regular and coach class air fares have stayed at their present level or decreased
slightly in spite of a rising cost-of~living index.
This has been made possible by the introduction of more efficient equipment, and it is predicted that
the introduction of turbine-powered aircraft (both
turbo-jet and turbo-prop) will allow fares to be
lowered even more.
In summary the author of this report holds that the superior characteristics of air
transportation over other transportation media have accounted directly for its rapid growth in the past and will account for its growth in the future.
It is pointed out that the other transportation media have pretty well used up their potential for improving efficiency, whereas the aeroplane is still in the process of technological development. Thus, it is forecast that the pas t rapid growth trend in air transportation will continue, although it is noted that even this forecast will probably prove to be too cons ervative.
Table 12 below (reproduced from the report) shows the actual growth of u.s. domestic air trave l from 1930 to 1955 and g i ves the f ore cast for the period 1955 to 1970. 106 TABLE 12
ACTUAL AND FUTURE GROWTH IN U.S. DOMESTIC AIR TRANSPORTATION
Scheduled Passenger-miles, Growth Increment and Growth Rate
Scheduled Growth Growth passenger-miles Increment rate Year (millions) (millions) %
1930 85 1931 107 22 25.9 1932 127 20 18.7 1933 175 48 37.8 1934 190 15 8.6 1935 316 126 66.3 ~936 439 123 36.1 1937 412 -27 -6.6 1938 480 68 16.5 1939 683 203 42.3 1940 1,052 369 54.0 1941 1,385 333 31.7 1942 1,418 33 2.4 1943 1,634 216 15.2 1944 2,178 544 33.3 ....:! 1945 3,362 1,184 54·4
1955 19,170 2,400 14.3 1956 21,770 2,600 13.6 1957 24,570 2,800 12.9 1958 27,570 3,000 12.2 1959 30,750 3,180 11.5 1960 34,100 3,350 10.9 1961 37,590 3,490 10.2 E-l 1962 41,220 3,630 9.7 ifJ. < 1963 44,980 3,760 9.1 ü 1964 48,850 3,870 8.6 ~ 1965 52,820 3,970 8.1 0 ~ 1966 56,880 4,060 7.7 1967 61,020 4,140 7.3 1968 65,230 4,210 6.9 1969 69,500 4,270 6.5 1970 73,820 4,320 6.2 107
It can be seen that the method used in this
forecast is an adjusted extrapolation of the trend
established by past air performance. The trend was
projected through on semi-logarithmic graph paper,
thus indicating changes in the rate of growth. The
declining rate of growth through to 1970 reflects
the opinion of the forecast's author regarding the
inability of air transportation to maintain a steady
growth rate through the forecast period.
It seems that the method used in this
forecast was more a reaction against methods used
in previous forecasts, or rather, against the results
achieved by previous forecasts, than the outcome of a
serious investigation into forecasting methods. So many factors are ignored in formulating this forecast that, in effect, it represents only a very narrow
examination of each of severa! related industries with no regard whatsoever being paid to the relationship of these industries to each other or to the whole economie system of which they form a part.
Comparison of Forecasts and Actual Traffic Results
Presented below is Table 13 in which the four forecasts are compared with actual traffic results.
This comparison is made in order that the relationship between each forecast and the actual result may be seen and in order that the relationship among the forecasts themselves may be noted. 108
TABLE 13
COMPARISON OF FORECASTS u.s. Domestic Traffic
B I L L I 0 N P A S S E N G E R M I L EE F 0 R E C A S T S Cana- c - w % PNYA % CAA % dair % YEAR 1944 error 1950 error 1955 error 1956 error ACTUAL*
1950 7.00 -13.5 7.30 -8.9 7.95 1955 11.07 -84.0 18.80+ -8.3 20.37 1956 11.70+~103.9 20.70+ -15.2 21.78 -9 .. 5 23.86 1957 12.50+~128.1 22.90+ -24.5 24.57 -16.1 28.52 1958 13.30+ 25.00+ 27.57 1959 13.80+ 27.20+ 30.75 1960 14.33 29.00 34.10 1965 16.71 37.00 52.82 1970 19.33 73.82
* Source: Facts and Figures - 1958, Air Transport Association of America, Washington, D.C.
+ Interpo1ated 109
The above table shows that all the forecasts
discussed here have been pessimistic. An earlier
Port of New York Authority forecast made in 1944 had
been unduly optimistic, having forecast 30 billion
revenue passenger-miles to be performed in 1955,
against the 20.37 billion actually performed. By
1950, the fact that this 1944 forecast was overly~
optimistic had become apparent and it probably was a
factor in the subsequent more pessimistic
prognostications.
Conclusions
Examination of the forecasts reveals a great
variation in magnitude. eompare the Port of New
York Authority's 1950 forecast with Canadair's
forecast. The former gives a figure of 19.33
billion revenue passenger-miles to be 'performed in
1970, whilst the latter gives that year's figure as
73.82 billion- a difference of over 53 billion ! However, even more significant is the fact
that the PNYA's forecast for 1970 was exceeded by the result for 1955. (1970 forecast - 19.33 billion:
1955 actual - 20.37 billion). That i s , actual performance exceeded f orecast performance fifteen years early ! llO
In comparing the CAA forecast and the Canadair
forecast, it is to be noted that the CAA forecast is
consistently lower than Canadair's, and that Canadair's
forecast itself appears to be conservative, as was
suggested by its author when he made the forecast.
Whilst all the forecasts discussed here have
proved to be pessimistic in the light of subsequent
performance, the PNYA and CAA forecasts are obviously
more pessimistic than the Canadair forecast. In terms
of percentage error in relation to the distance in
time from the date of forecast formulation to the
forecast year, the Curtiss-Wright report compares
favourably with the others, since, after the six
year period from the date of forecast formulation
(1944) to the forecast year (1950), there was on1y
a 13.5 percentage error.
All the forecasts discussed here appear, so
far, to have been rather inaccurate. Viewed from one air1i nefs standpoint "inaccurate" cou1d mean that, had that air1ine correctly forecast its share of the total market, its aircraft purchasing plans based upon the industry f orecast wou1d have resu1ted in that air1ine having either too few or t oo many aircraf t to economLca11y satisfy the demand for air transportation. 111
Three of the four forecasts noted the
existence of a correlation between inter-city trave1
and GNP. One of these three forecasts noted the
existence of a correlation between per capita inter
city trave1 and national income. However, ail three
of these forecasts dea1t with inter~city travel in
total terms, and did not attempt to discover if a
correlation existed between GNP and per capita inter
city travel by a specifie medium of transportation.
In all cases discussed here, the forecasters tended to be over~cautious, probably as a result of their subjective feeling that the rate of growth in u.s. domestic air trave1, current at the time the forecast was made, could not be maintained.
This subjective feeling may have been due to a preoccupation with a vaguely conceived sense of
"saturation." The forecasters appeared to have reasoned that, somewhere in the future, the saturation point for air travel would be reached.
However, they made no attempt to discover what this saturation point was.
It would have been better if the forecasters had attempted to discover the "saturation point".
For example, it would have been possible to define the 112
saturation point for air travel (and for travel by
any other medium) for any particular year by examining the past trend of per capita air travel, by relating
this per capita trend to GNP, and, with the use of a
forecast of GNP and population, arriving at a
"saturation" figure of air .passenger-miles for the year in question.
This figure woald have provided the upper limit for the forecast, for that particular year, and if it turned out, for example, that this figure would be achieved if the current rate of growth were maintained, there would have been no reason for the subjectively inspired decreases in the rates of growth to have beeri introduced into the forecasts, decreases which resulted in under-estimates of varying degrees.
Another method by which the saturation level could have been determined would have entailed considerably more research effort and expense than the forecas ters were probably willing or able to expend. This method is commonly known as the "cell" method and involves the following procedure:
The population for whom it is intended to formulate a travel forecast i s , for the purpose of the research, divided into "cells," the types of cells used, for example, being:
1) a ge (e.g. under 20, 20 to 30, 30 to 50,
50 to 65, ove r 65) 113
2) income level (e.g. under $3000
per annum, $3000 to $6000, etc.)
Scientifically selected samples of the population presently within each cell would be interviewed about their travel habits of the past and their expected travel habits for the future. From this sample would be determined the travel habits of the sum of all individuals within each cel!, for example, within the cell "age 20 to
30 years, income $3000 to $6000 per annum". Since the number of individuals within each cell would also be known, the per capita travel per cell could be calculated.
A figure for the total travel of the population by all media of transportation could thus, be calculated, as could also be a figure for the total travel of the population by each medium of transportation. Where statistics on actual travel are available (e.g. statistics on the number of inter-city passenger-mLles performed), the calculated figures based upon the cel! sampling could be checked against the actually achieved figures, and the accuracy of the sampling method thereby tested. 114
Using detailed forecasts of national income and its distribution, and detailed forecasts of population and its distribution by age group, the size of the population within each cell in the forecast period can be calculated.
By applying the previously determined per capita travel habits to the appropriate forecast cell, it is possible to forecast the total amount of travel to be undertaken by the individuals within that cell, and, therefore, to forecast the total amount of travel to be undertaken by the whole population. Separate forecasts of the amount of travel to be undertaken by each medium of transportation can also be constructed in the same way.
However, the forecasts for each transportation medium would have to be adjusted, where necessary, to take into account expected changes in the competitive position of one or more of the media. For example, it would be unrealistic to forecast a growth in trans~ continental passenger railroad traffic merely on the basis of a forecast growth in the population of a cell which in the past has generated much of such traffic, if it is known that the railroads intend curtailing or eliminating such passenger services. 115
By the use of this cell method it is possible to forecast the total travel of the population, and the share which each medium of transportation is expected to obtain.
Since a forecast formulated by the use of such a method assumes that changes in the size of an individual cell will result in changes in the total volume of travel generated by that cell, and in changes in the total volume of travel by medium of transportation {subject to the adjustments previously noted) generated by that cell, it follows that such a forecast also represents the "saturation" level, for total travel and for each transportation medium. CHAPTER VI
FORECAST OF REVENUE PASSENGER KILOMETRES TO BE PERFORMED BY SCHEDULED AIRLINES IN BRAZIL 116
FORECAST OF REVENUE PASSENGER KILOMETRES TO BE PERFORMED BY SCHEDULED AIRLINES IN BRAZIL
CHAPTER VI
Under ideal conditions, conditions under which all the required data were available, I would construct a forecast of scheduled airline passenger traffic in Brazil {or, indeed, in any country) in the following manner:
1) I would obtain historical
statistical data on inter-
city passenger traffic, by
medium of transportation
(i.e. automobile, bus, rail and air).
2) Also obtained would be past
statistics of gross national
product {in constant cruzeiros),
and an authoritative forecast
of future GNP.
3) The geographie, cultural and
economie structure of the
country would be studied, in
order that the e xisting
situation in which the various
transportation media operate
would be fully understood, and
in order that the significance
to transportation of expected
changes in this structure could
b e adequately assessed. 117
4) The degree of correlation
between GNP and total inter
city passenger travel would
be determined, and, if the
degree of correlation were
sufficiently high, inter-city
passenger travel would be
projected through the forecast
period, based upon the forecast
of GNP.
5) The degree of correlation between
GNP and each individual
transportation mediumts traffic
would also be determined, and,
in these cases where a high
degree of correlation existed
(probably in inter-city bus,
automobile and air travel,
in Brazil), a forecast of the
individual mediumts performance
would be made. Any one of these
forecasts would be modified
should the study of the countryts
structure have revealed an
existing or impending situation
likely to have a significant
effect on the performance of
a particular medium of
transportation (e.g. the 118
development of rich but remote
resources would be likely, over
a fairly short period of time,
to result in a significant
growth of air passenger traffic).
6) Forecasts of the performance of
those transportation media which
have a low degree of correlation
with GNP would be made on the
basis of an individual study of
their past performances and of
the factors which appear to be
closely related to these
performances.
7) The sum of the individually
constructed forecasts would
then be compared with the directly
constructed forecast of total
inter-city traffic. Should there
be any significant difference
between the two totals both sets
of forecasts would be re-axamined
and adjustments made where
considered necessary. 119
8) The share of the total transportation market which has
accrued to each transportation
medium in the past would also
be examined, in order to
determine if a trend exists in
the share obtained by any of
the media, Should it be determined
that a definite trend existed for
one or more of the transportation
media, this trend would be
projected through the forecast
period, and a forecast for the particular medium of transportation
obtained by applying the projected
share to a forecast of total
transportation derived, for
example, through a correlation of
GNP and transportation, as
described in 4) above. Unfortunately,these techniques cannot be used since, in Brazil there are not, to my knowledge, any published statistics on the inter-city passenger kilometres performed by railroads, by autobuses, or by private automobiles. The International Civil
Aviation Organization (ICAO), however, does publish statistics on Brazilian airline pâssenger-kilometres performed. 120
An attempt was made to estimate the number of private automobile inter-city passenger~ kilometres performed in 1957 in Brazil. A description of the method used and the estimated figure is contained in Appendix II to this
Chapter. However, it is considered that the estimate. so derived cannot be considered reliable, since the assumptions used cannot be checked.
Therefore, the attempt to estimate the share of total inter-city passenger-kilometres which each transportation medium has obtained in the past was abandoned.
In addition to the difficulty posed by lack of total inter-city passenger-kilometre statistics, there is not, to my knowledge, any forecast of GNP prepared or published by any official or un-official Brazilian organization.
Thus, it was not possible to forecast Brazilian airline traffic on the basis of a possible correlation between total traffic and GNP.
Being faced with this lack of statistical data, I therefore considered it desirable to locate a fairly predictable economie indicator to which
Brazilian air passenger travel appeared to be directly and highly correlated. 121
It was determined that the Index of
Industrial Production for Brazil was such an indicator. The indices were obtained from the monthly statistical bulletins published by the
United Nations. Statistics on air passenger kilometres performed were obtained from the digest of statistics published by ICAO.
The degree of correlation which was ca1cu1ated to exist in Brazi1 between the index of industrial production and air passenger trave1, over the period 1950 to 1958 inclusive, was 0.989. In order to determine whether such a high degree of correlation was typical of other countries in roughly the same stage of deve1opment as Brazil, or whether Brazi1ts case was unique, the same correlation coefficient (index of industria1 production to air passenger-kilometres performed) was a1so ca1cu1ated for Argentina, India and Mexico. The res u1ts were, respecti ve1y,
0.9707, 0.9946 and 0.9807. The ca1culations resulting in these coefficients, inc1uding the
Brazi1ian coefficient, are shawn in Appendix I to this Chapter. 122
In the 1ight of the correlation coefficients ca1cu1ated for the other countries, it appears that the Brazilian case is not unique, and it therefore
seems reasonable to accept the index of industria1 production as the basis for formulating a forecast
of Brazilian air passenger-kilometres to be performed.
The problem of forecasting passenger kilometres thus hinges upon a forecast of the index of industria1 production.
Although, as previously noted, there appears to be no forecast of Brazilian GNP avai1ab1e from Brazilian sources, there is a forecast of this factor contained in a United
Nations publication dealing with the economie development of Brazil • This forecast of GNP
(in constant cruzeiros) is made up to the year 1962.
Beyond that year, for the purpose of this thesis, the forecast has been projected through to the year 1965. On semi-logarithmic graph paper, the
United Nations forecast from 1957 to 1962 is a straight line, and this straight line has been continued through to 1965, as is shown on Graph 1 below. 123
GRAPH l .t BRAZILIAN GNP (IN 1952 CRUZEIROS) ACTUAL TO 1956: UNITED NATIONS FORECAST 1956 TO 1962: STRAIGHT PROJECTION 1962 TO 1965
10 0( 9 rn 8:oo ~700 ·~6 00 4) - ~-· oq ~5 ... -- --· -- ~ .... -- - u400 -·- ~ ~ A., V) Zo ~ - C!l r-i3 on --- Ci-! 0
~2·~ 0 r-t ""r-t •.-l j:Q 10 0 19 51 51 ~2 ">3 4 5' 5o 57 58 . 5} oo n 62 63 6~ 6
SOURCE: Analyses and Projections of Economie Development, II THE Economie Development of Brazil (New York, United Nations, 1956) p.31
It was calculated that there exists, as was to be expected, a high degree of correlation between Brazilian GNP and the index of industrial production. The coefficient of correlation in this case is 0.9981, and the calculations which determined this are contained in Appendix III to this Chapter. 124
Having determined that this high degree of correlation existed between GNP and the index of industrial production, and having available a forecast of GNP, it was possible to forecast the index of industrial production through to 1965.
The figures of the forecast of the index of industrial production were read off the graph contained in the previously mentioned Appendix
III to this Chapter. These figures and the GNP figures are given in Table 14 below:
TABLE 14
GNP&INDEX OF INDUSTRIAL PRODUCTION 1950 - 1965
GNP (billions of Index of Years 1962 cruzeiros ) Industrial Productior
1950 324.1 81 1951 346.5 88 1952 360.9 96 1953 376.1 lOO 1954 409.2 108 1955 418.1 114 1956 428.0 122 1957 448.2 126 1958 469.6 132 1959 491.9 139 1960 514. 8, 148 1961 538.2 155 1962 564.7 164 1963 590.0 173• 1964 618.0 182 1965 643.0 191 1 125
Having derived a forecast of the index of industrial production through to 1965, it was thus possible to read off the forecast for scheduled revenue passenger-kilometres to be performed by Brazilian airlines up to 1965 from the graph showing the relationship between passenger kilometres performed and the index of industria1 production contained in Appendix I to this Chapter.
The forecast so derived, with the re1ated indices of industria1 production, is given below:-
TABLE 15
INDEX OF INDUSTRIAL PRODUCTION AND SCHEDULED REVENUE PASSENGER KILOMETRES 1959 - 1965
Schedu1ed Revenue Year Industria1 Production Passenger-Ki1ometres (hundred mi11ionl
1959 139 25.9 1960 148 28.5 1961 155 30.5 1962 164 33.2 1963 173 35.8 1964 18,3 38.5 1965 191 41.2 126
It is, in my opinion, impossible at this time to estimate, with any degree of authority, whether the forecast for the year 1965 represents the attainment of a "saturation point" for the Brazilian air passenger industry, since there does not exist, to my knowledge, any really satisfactory standard against which to measure the forecast, in terms of saturation. In this regard, I examined the comparable statistics for United States airline traffic for the year 1958, and determined that U.S. airlines performed 291 passenger-kilometres per head of population in that year. The per capita figure for
Brazilian airlines for the same year was 41, whilst, assuming that the trend of Brazilian population increase is the same as that experienced over the last thirty years, the per capita number of passenger-kilometres to be performed in 1965, according to the forecast, would only be 56.
However, although these figures would seem to indicate a great un-realised potential in
Brazil, when the levels of income in the United
States and in Brazil are taken into account, a somewhat different interpretation could be put upon such figures. In the United States in 1958, the per capita national income in current dollars was $2,072. 127
In Brazil in the same year the per capita national
income was approximately $100 (current cruzeiros
converted to dollars at the 1958 free rate of
exchange).
Therefore, in terms of income, the
Brazilian airlines should have performed only
approximately one twentieth of the number of per
capita passenger-kilometres as the United States
airlines in 1958, whereas, in fact, Brazilian
airlines actually performed approximately one
seventh the number of per capita passenger
kilometres performed by the U.S. airlines in that
year.
The fact that per capita air passenger
traffic in Brazil is out of proportion to Brazilian
per capita income, measured in terms of U.S. per
capita air passenger traffic and per capita income,
indicates that air transportation in Brazil must
possess special advantages over other media of
transportation. These special advantages must
derive from Brazil's geographie characteristics,
from the state of development of the surface transportation facilities, and from the ability
of the aeroplane to overcome natural obstacle to 128
surface transportation. This conclusion is borne
out to a certain extent by the comparable per capita figures for Colombia, Mexico and Brazil shown in
Tablel6 below:
TABLE 16
1958 per capita 1958 per capita* Country air passenger-kilometres National Income
Colombia 46 $116 Mexico 60 $252 Brazil 41 $100
* u.s. dollars converted from local currencies at 1958 free rates of exchange. Appendix I to Chapter VI
Calculations of the coefficients of correlation:
Index of Industrial Production/ Air Passenger-kilometres
Brazil Argent ina India Mexico 129
B R A Z I L
Scheduled Revenue Passenger Index of kilometres Theoretical Industrial (hundred Regression Year Production million) Values (N) x y Xx Y x2 y2 yc Y-Yc d2
1950(1) 81 8.5 689 6561 72.25 8.8 -0.3 -0.09 1951(2) 88 12.4 1091 7744 153.76 10.8 +1.6 +2.56 1952(3) 96 12.8 1229 9216 163.84 13.2 -0.4 -0.16 1953(4) 100 14.8 1480 10000 219.04 14.4 +0.4 +0.16 1954(5) 108 15.9 1717 11664 252.81 16.7- -o.8 -0.64 1955(6) 114 16.8 1915 12996 282.24 18.4. -1.6 -2.56 1956(7) 122 18.8 2294 14884 353.44 20.8. -2.0 -4.00 tl957(8~ 126 22.9 2885 15876 524.41 22.0 +0.9 +0.81 ~958(9 132 25.9 3419 17424 670.81 23.8 +2.1 +4.41 967 148.8 16719 106365 2692.60 +0.49
SOURCES: Indices of Industrial Production; Monthly Bulletins of Statistics, Statistical Office of the United Nations, New York
Scheduled Revenue Passenger-kilometres; Digest of Statistics No. 75 (1947-1958), International Civil Aviation Organization, Montreal 130
B R A Z I L
Line of Regression (!east squares)
Equation (I) :( (Y) = Na + b :( (X) Equation (II) :( (XY) = a ~(x)+b :( (X2)
(I) 148.8 = 9a + 967b (II) 16719 = 967a + 106365b 15981 967a + 103856b Equation(I) x 107.4 Subtract 738 = 2509b
The re fore b + 0.294
Substituting the value of b in Equation (I)
148.8 = 9a + 967 (0.294) 148.8 9a +284.3 9a = 148.8 - 284.3 9a = - 135.5 Therefore a = ~ 15.05
Line of Regression Y = a + b (X)
Therefore Y = - 15.05 + 0.294(X) 131
BR A Z I L
Standard Err or of Estima te {SI;)
"'
Sy = ~ (d2) ) N
Sy = 0142 J 9 Sy = Jo.oss
Therefore Sy = 0,074 132
B R A Z 1 L
Standard Deviation (cry) cry : ;~!2l - t:y 12
cry =:: 2623 2 J 9 (1~8.8) cry =:: )299.2 273.2
cry =:: J26.o
Therefore o-:r =:: S.l 133
B R A Z I L
Coefficient of Correlation (r)
r ... )1 ~cry
r o!ss )1 26.00
r = ,j1 0.02114
r = )0.979
Therefore r = 0.989 134
BRAZIL
PASSENGER KILOMETRES - INDEX OF INDUSTRIAL PRODUCTION
Index of Industrial Production x 135
A R G E N T I N A
Scheduled Revenue Passenger Index of Kilometres Theoretical Industrial (hundred Regression Year Production million) Valuee (N) x y XxX x2 y2 Yc Y-Yc d2
~950(1) 105 2.64 277.2 11025 6.97 2.90 -0.26 -0.0676 ~951(2) 108 3.05 329.4 11664 9.30 3.28 -0.23 .,;.0.0529 ~952(3) 101 2.46 248.5 10201 6.05 2.40 +0.06 -0.036 ~953(4) lOO 3.17 317.0 10000 10.05 2.29 +0.88 +0.7744 1954(5) 109 3.42 372.8 11881 11.70 3.42 0 0 1955(6) 118 3.71 437.8 13924 13.76 4.55 -0.84 -0.7056 1956(7) 117 4.39 513.6 13689 19.27 4.42 -0.03 -0.0009 ~95718~ 121 5.15 623.2 14641 26.52 4.92 +0.23 +0.0529 :!_958 9 125 6.29 786.3 15625 39. 56 1 5.42 +0.87 +0.7596 1004 34.28 3905.8 112650 143.18 +0.7239
SOURCES: Indices of Industrial Production; Monthly Bulletins of Statistics, Statistical Office of the United Nations, New York
Scheduled Revenue Passenger-kilometres; Digest of Statistics No. 75 (1947-1958), International Civil Aviation Organization, Montreal 136
A R G E N T I N A
Line of Regression (!east squares)
Equation (I) ~(Y)= Na+ b~(X) Equation (II) ~ (XY)= a ~(X)+b ~ (X2)
(I) 34.28 = 9a + 1004b
(II) 3905.8 1004a + 112650b 382:1.1. = lOO :ta + 11200lb Equation(!) x 111.555 Subtract 81.7 649b
Therefore b = + 0.12588
Substituting the value of b in Equation (I)
34.28 9a + 1004 (0.12588) 34.28 = 9a + 126.38 9a = 34.28 126.38 9a = -92.70 Therefore =a=====~=l=0=·~3
Line of Regression Y a + b (X)
The re fore Y = -10.3 + 0.12588 (X) 137
A R G E N T I N A
Standard Error of Estimate (Sy)
Sy =
Sy = 0,7239 9
Sy = jo.o8043
Therefore Sy = 0,2836 138
A R G E N T I N A
Standard Deviation (cry)
=
=
= 15.9088 14.5153
= ;1.3935
Therefo~e = 1.1803 139
A R G E N T I N A
Coefficient of Correlation (r)
r
r = (0.2836)~ (1.1803)
r 1 0,08043 1.3935
r 0.057717
r = J 0.942283
Therefore r 0.9707 140
ARGENTIN A
PASSENGER KILOMETRES - INDEX OF INDUSTRIAL PRODUCTION
y
Index of I ndustrial Production x 141
IN DI A
Scheduled Revenue Passenger- Index kilometres il'heoretical Industrial (hundred R.egression Year Production million) Values (N) x y Xx Y x2 y2 Yc Y-Yc d2
1951(1 95 4.14 393.3 9025 17.14 3.55 +0.59 +0.35 1952(2 98 3.89 381.2 9604 15.13 3.89 0 0 1953(3 lOO 3.86 368.0 10000 14.90 4.12 -0.26 -o.o68 1954(4 107 4.40 457.6 11449 19.36 4.91 -0.51 -0.26 1955(5 116 5.15 597.4 13456 26.52 5.93 ..o.78 -0.61 1956(6 126 6.79 855.5 15876 46.10 7.16 -0.37 -0.14 1957(7~ 130 7.82 1016.6 16900 61.15 7.50 +0.32 +0.10 1958(8 132 8.66 1143.1 17424 74.99 7.74 +0.92 +0.85 904 44.71 5230.7103734 275.29 +0.272
SOURCES: Indices of Industrial Production; Monthly Bulletins of Statistics, Statistical Office of the United Nations, New York
Scheduled Revenue Passenger-kilometreSJ Digest of Statistics No. 75 (1947-1958) International Civil Aviation Organization, Montreal 142
I N D I A
Line of Regression (1east squares)
Equation (I) ~(Y) = Na + b ~(X) Equation (II) ~(XY) = a ~(X) +b~x2)
(I) 44.71 Ba + 904b (II) 5230.7 = 904a + 103734b so 52. 2 204a + 1021S2b Equation (I) x _ll3 Subtract 17B.5 = 15B2b
therefore b = + 0.112B3
Substituting the value of b in Equation (I)
44.71 = Ba + 904 (0.112B3) 44.71 = Ba + 101.99 Ba = 44.71 101.99 Ba -572B 'lherefore a = ~7.16
Line of Regression Y a + b (X)
therefore Y = - 7.16 + 0,112B3 (X) 143
I N D I A
Standard Error of Estimate (Sy)
Sy
Sy 0.272 8
Sy =~
Therefore Sy = 0.05831 144
I N D I A
Standard Deviation (~y)
Therefore = 1.7825 145 I ND I A Coefficient of Correlation (r) r = 1 - ayz-~ 2 r 1 ( 0. 05831) (1.7825) 2 r = 1 0.034 3.17712 r F Therefore r = 0.9946 146 PASSENGER KILOMETRES - INDEX OF INDUSTRIAL PRODUCTION 9o do Ùo Index of Industrial Production x 147 M E X I C 0 Scheduled Revenue Passenger- Index of kilometres rrheoretical Industrial (hundred ~egression Year Production million) ~alues (N) x y XxY x2 y2 Yc Y-Yc D2 1950(1) 93 6.25 581.25 8649 39.06 8.90 -2.65 -7.05 1951(2) 99 8.71 862.29 9801 75.86 10.05 -1.34 -1.79 1952(3) 99 9.80 970.20 9801 96.04 10.05 -0.25 -o.o6 1953(4) lOO 13.45 1345.00 10000 180.09 10.25 +3.20 +10.24 1954(5) 107 13.90 1487.30 11449 193.21 11.60 +2.30 +5.29 1955(6) 119 13.66 1625.54 14161 186.60 13.90 -0.24 -o.o6 1956(7l 131 15.13 1982.03 17161 228.92 16.20 -1.07 -1.14 1957(8 139 17.55 2439.45 19321 308.00 17.75 -0.20 -0.04 1958(9 147 19.4S 2859.1_5_ 2160_2_ _3_18 ._30 1_2_.28 +0.17 +0.03 1034 117.90 14152.21 121952 1686.08 +5.42 SOURCES: Indices of Industrial Production; Monthly Bulletins of Statistics, Statistical Office of the United Nations, New York Scheduled Revenue Passenger-kilometres; Digest of Statistics No. 75 (1947-1958), International Civil Aviation Organization, Montreal 148 M E X I C 0 Line of Regression (least squares) Equation (I) ~(Y) = Na + b ~(X) Equation (II) ~(XY)= a~ (X) + b~ (X2 ) (I) 117.90 = 9a + 1034b 14152.21 = 1034a+ 121952b 13544.35 = 1034a+ 118786b Equation(!) X 114.8E Subtract 607.86 = 3166b Therefore b = + 0.192 Substituting the value of b in Equation (I) 117.90 = 9a + 1034(0,192) 117.90 = 9a + 198.53 9a = 117.90 - 198.53 9a = -80.63 Therefore a = -8.9 59 Line of Regression Y = a + b(X) Therefore Y = -8.959 + O,l92(X) 149 MEXICO Standard Error of Estimate (Sy) Sy ~/"'(tl Sy F Sy Jo.6ozz Therefore Sy = 0.776 150 MEXICO Standard Deviation (~y) = = 187.34 171.61 cry ~15.73 Therefore av = 3.966 151 MEXICO Coefficient of Correlation (r) 2 r - ~ r = r 0,6022 15.73 r 0.0383 r = J 0.9él7 Therefore r 0.9807 152 MEXICO PASSENGER KILOMETRES - INDEX OF INDUSTRIAL PRODUCTION Index of Industrial Production x Appendix II to Chapter VI Estimate of the number of private automobile inter-city passenger kilometres performed in Brazil in 1957 153 Estimate of the number of private automobile inter-city passenger-kilometres performed in Brazil in 1957 In order to estimate the number of private automobile inter-city passenger-kilometres performed in Brazil in the year 1957, the ratio of the figure for the same traffic in the United States for that year to the number of surfaced miles of road in the u.s. was first examined. The number of U.S. passenger-miles so 41 performed in 1957 was 637,755,000 • The number of miles of surfaced road in the United States in 1957 42 was 2,411,000 • Therefore, the ratio of automobile inter-city passenger-miles to miles of surfaced road was 264,518 to one. In Brazil in 1957, the total length of paved roads was 5, 666 kilometres 43• Therefore, if it is assumed that the ratio of private automobile inter-city passenger travel to the length of paved roads was the same in 1957 in Brazil as it was in the United States in that year, then the number of private automobile inter-city passenger-kilometres performed in Brazil in 1957 would have been 5,666 times 264,518, or 1,498,758,988. 41- u.s. Department of Commerce, Statistical Abstract of the United States - 1959, p. 565 42- !!2i.ç!, p.548 43- u.s. Department of Commerce, Comparative Statistics on the Latin American Repub1ic - 1957, p.5 154 Another ratio was also examined in an attempt to estimate the number of private automobile inter-city passenger-kilometres performed in Brazil in 1957. This ratio was that existing between the number of private automobile inter- city passenger-miles performed in the United States in 1957 to the number of registered private automobiles existing in that year. For the United States this ratio is 11,451 passenger~ miles (or 18,422 passenger-kilometres} per registered automobile. Assuming that the same ratio existed in Brazil in that year, the number of automobile inter-city passenger~kilometres would have been 7,293,435,598 (18,422 kms. x 395,909 registered automobiles). Using this estimated figure in order to check the previously determined ratio of automobile inter-city passenger-kilometres performed per kilometre of paved road gives a result of 1,287,228 to one (7,293,435,595 divided by the number of kilometres of paved road- 5,666), as against the ratio of 264,518 to one previously used. 155 Comparing these ratios to each other (i.e. comparing the ratio of 1,287,228 to one, to a ratio of 264,518 to one), it would appear either that the use of the "inter-city passenger travel to number of automobiles" ratio method is impractical for estimating Brazilian automobile inter-city travel, or that the paved roads of Brazil carry a much greater volume of automobile traffic than do the paved roads of the United States. This second possibility,surprising as it may initially seem, appears to be supported by the figures showing the number of automobiles per kilometre of paved road in each country for the year 1957. In the United States in 1957 there were 44 55,693,000 registered automobiles • The surfaced road mileage was, as previously noted, 2,411,000. Therefore, there were 23.1 registered automobiles per paved road mile, or 14.3 automobiles per paved road kilometre. In Brazil in the same year there were 395,P09 registered automobiles and 5,666 kilometres of paved road, thus indicating that there were 69 automobiles per paved road kilometre - just less than five times the. automobile-to-available- road density that existed in the United States. 44- u.s. Department of Commerce, Statistical Abstract of the United States - 1959, p.560 156 And, as previously noted, assuming that there were performed 7,293,435,598 passenger kilometres in automobile inter-city travel in Brazil in 1957, the ratia of automobile inter city travel to kilometres of paved road would be 1,287,228 to one - just less than five times the ratio of the same two factors in the United States (264,518 to one). Therefore, in view of the closeness of the ratio of automobiles per road kilometre to the ratio of inter~city passenger-travel per paved road mile or kilometre, it is f elt reasonable to assume that the number of private automobile inter-city passenger-kilometres performed in Brazil in 1957 was approximately 7 billion. Appendix III to Chapter VI Calculation of the coefficient of correlation between Brazilian Gross National Product and the Index of Industrial Production 157 B R A Z I L GNP (billions Index of Theoretical of 1952 Industrial Regression Year Cruzeiros) Production Values (N) x y Xx Y x2 y2 Yc Y-Yc d2 1950(1) 324.1 81 26252.1 105040.8 6561 82.5 -1.5 -2.25 1951(2) 346.5 88 30492.0 121062.3 7744 90.0 -2.0 -4.0 1952(3) 360.9 96 34646.4 130248.8 9216 94.5 +1.5 +2.25 1953(4) 376.1 100 37610.0 141451.2 10000 100 0 0 1954(5) 409.2 108 44193.6 167444.6 11664 111.5 -3.5 -12.25 1955(6) 418.1 114 47663.4 174807.6 12996 114 0 0 1956(7~ 428.0 122 52216.0 183184.0 14884 117.5 +4.5 +20.25 1957(8 448.2 126 56473.2 200883.2 15876 124.3 +1.7 +2.89 3111.1 835 329546.7 1224122.5 88941 +6.89 SOURCES: GNP; Analyses and Projections of Economie Development, II; The Economie Development of Brazil, Department of Economie & Social Affairs of the United Nations, New York, 1956 Indices of Industrial Production; Monthly Bulletins of St~tistics, Statistical Office of the United Nations, New York 158 B R A Z I L Line of Regression (least squares) Equation (I) ~(Y) = Na + b ~(X) 2 Equation (II) ~(XY) a ~ (X) +b ~(X ) (I) 835 = 8a + 3lllb (II)329547 = 3llla + 1224123b 324711 Jllla + 1209790b Equation(I)X388.875 Subtract 14333b Therefore b = +0.3374 Substituting the value of b in Equation (I) 835 = 835 8a = 8a = a = Line of Regression Y = a + b (X) Therefore Y= -26.825 + 0.3374 (X) 159 B R A Z I L Standard Error of Estimate (Sy) Sy Sy = .Q.d9. 8 Sy = J0.86125 160 B R A Z I L Standard Deviation (cry) 88941 8 _/ 11117.625 - 10894.141 = J 223.5 161 B R A Z I L Coefficient of Correlation (r) r s 2 ~2 r = 1 0,86125 223.5 r = 0,00385 r = Therefore r 0, 9981 ======~=== 162 BRAZIL INDEX OF INDUSTRIAL PRODUCTION - GROSS NATIONAL PRODUCT - ---·--r~~~~:-:""'T~~"'":"""..,...... ~==_,.,"'":"""..,...... ~~..,...-"""'T"'"..,...-~~~""!""r"~~,...,..~"""':7'=1- y 190 180 170 160 150 ~f- -TL_L~- ~,~,~ r -1·-~--l,_ . - - -~1--T-+- - t: 140 0 ·1"1 .j.) ::s(.) 'C f 130 ~ ~ 111 ·1"1 "~'""····· .• :,:!, --·~~ ,, .. ,.IJ . B4:.~.-~ ..l ...0 j. 1:•.•~ :, ~"' 120 :·i: ::· ··• : ~ s: B~ r:)~ciijG1,i~~M~:~\~:+J - i±-(! - ': IJ" rr's '~' .r_ 'C·= ~ ~;:; :::; .:: :;:: :::.,;'. ·(.· -:de'~ ~~H~--Hs:, :.;Hr: j: ·:· _',. ~-H-:: ~ ;_:t-·p: i iT c-8rt'l~:. + i~ 8.c:c-f,.,..., ·::tee: .,.J .. ·,-,1~-54,-.,.--,.- , .. ,~,.,T .. l .. ,·;:r- .. ,., - -.-::1 "'i'"' .,,'", .., . ~. .. --, ,-, .. -·:·. '·ir'-r,,.. ,,.,.,. ..,. :.--;.. 100 %~~ ~~ m~~J ~~-~~~~:l~t~ ~ill~tJi:~~r-f-t~t~~ID~JJ~--~;l 1 ....•. - ;r:; ·- ·- :-~ntJ~f ,f-. ·'-l·... ··,.. .·r'-.,1 ..L ::-.- ·. ·1-:.-. . '-.-. ·.• '.L 1... -...•. -- .. -,r·,-... ,. ,~-,-,, .. 1. - ,.-, -.. -,J•. , .. ,.·-.!-. -,-! - .·-Jl :.. ,.. , ..• J.., ... ,._.. 1 ,, ... 1L .. " :. -- 1, ...._ :, .. ,~.! .,,•.- ,.-.+.-.t.-,·'!'.'" .-l '.."' 1' -- -.1,, .. ''• . 90 80 ! 70 ·i - L · :--:·· i-· ·· i . GNP Billions of 1952 Cruzeiros x B I B L I 0 G R A P H Y 163 BIBLIOGRAPHY Aldrich, G.H., Hart, A.L., McGoldrick, J.D., Owen, W., & Sorrell, L.C., Air Traffic Forecast 1950-1980, (Port of New York Authority, New York, 1950). Almeida, M.O. de, Chermont, J.S., & Vasconcellos, D.A. de, Survey of the Brazilian Economy, (Brazilian Embassy, Washington, 1958). Bratt, E.C., Business Cycles and Forecasting, (Richard D. Irwin Inc., Chicago, 1958). Dean, J., Managerial Economies, (Prentice-Hall Inc., New York, 1951). Dewey, E.R., & Dakin, B.F., Cycles, The Science of Prediction, (Henry Holt and Co., New York, 1947). Larsson, K.H., A Critical Review of Earlier Forecas ts of Air Traffic and a New Approach, (Canadair Ltd., Montreal, 1956). Maisel, S.J., Fluctuation, Growth and Forecasting, (John Wiley and Sons Inc., New York, 1957). McDonald, B.A. & Drew, J.L., Air Transportation in the Immediate Post-War Period, (Curtiss-Wright Corp., Buffalo, 1944). National Council of Economies of Brazil, Analys is and Outlook for Brazilian Commercial Air Transportation, (Rio de Janeiro: 1959). Rickards, A.E. , A Statistical Analys is of the s cheduled airlines registered in the Americas (excluding Canada and the United States), and an estimate of their need for aircraft during the period 1959 to 1965 inclusive, (Canadair Ltd., Montreal, 1959). 164 United Nations, Analyses and Projections of Economie Development - II, The Economie Development of Brazil, (New York: 1956). U.S. Civil Aeronautics Administration, Civil Air Traffic Forecasts 1960-1965), (U.S. Government Printing Office, Washington, 1955). u.s. Department of Commerce, Basic Data on the Economy of Brazil, (U.S. Government Printing Office, Washington, 1959). u.s. Department of Commerce, Civil Aviation in Brazil, (u.s. Government Printing Office, Washington, 1956). u.s. Department of Commerce, Comparative Statistics on the Latin American Republics, (U.S. Government Printing Office, Washington, 1958). u.s. Department of Commerce, Economie Developments in Brazil in 1958, (U.S. Government Printing Office, Washington, 1959). PERIODICALS Anuario Estatistico do Brasil, (Annually) (Conselho Nacional de Estatistica, Rio de Janeiro, 1958). Aviation Daily, (American Aviation Publications Inc., Washington, February 1959 to-date). Digest of Statistics, (Annually) (International Civil Aviation Organization, Montreal, 1958). Flight Magazine (Monthly) (Iliffe & Sons Ltd., London, January 1959 to-date). International Financial Statistics, (Annually) (International Monetary Fund, Wa shington, 1958). Monthly Bulletin of Statistics, (United Nations, New York, January 1958 to~date). Statistical Abstract of the United States, (Annually) (u.s. Department of Commerce, Washington, 1959).