Distr, _
LIMITED =
E/CH.14/NAC/34
25. August I969 Originals ENGLISH
ECONOMIC COMISSIOF FOE AFRICA Seminar on Statistics of Prices
and Quanta, .
Addis .Ababa,,13-21 October I969
COUHOHY PRACTICES IK THE GAOHERIFG OF BATA OJT PRICES AHD QUASTTA AHD OEE CALCULATION OF RELATED
INDEX MJMBEES
M69-2261
COUNTS! PRACTICES IN TEE GATHERING OP DATA 01 PRICES AND QUANTA AND THE CALCULATION OF RELATED
INDEX NUMBERS
TABLE OP CONTENTS
. Page
INTRODUCTION 1
EGA QUESTIOMAIRE 3
AVAILABILITY OP L&TA , ... 4
PRODUCER PRICES AND PRICE INDICES 10
RETAIL OR CONSUMER PRICES AND PRICE INDICES 13 PRODUCTION STATISTICS AND INDICES OP PRODUCTION 19 EXTERNAL TRADE AID RELATED QUANTUM AND UMIT VALUE INDICES... 24 OUTPUT OP PRODUCERS. OP GOVERNMENT SERVICES 29
ANNEX: BASIC CHARACTERISTICS OF INDEX NUMBERS OP PRICES ASTD QUANTA AVAILABLE!
A. WHOLESALE PRICE,INDICES
B.. RETAIL OR CONSUMER PRICES INDICES
C. ' INDEX mJMBERS OP PRODUCTION
D. INDEX NUMBERS OP EXTERNAL TRADE AND OTHER INDICES
B/CH.14/M.G/34
COUNTRY PRACTICES. 1ST THE GATHERING OF DATA-OF PRICES AND QUANTA . ,. , AffiD .THE CALCULATION OF RELATED. INDEX NUMBERS
.INTRODUCTION: -■■■■-.. . ,; . . ■
.1. . This paper, is. intended. to. describe briefly the main features of national practices in the collection of data on prices and quantities and in the calculation of related index numbers, in countries of the
■African .region* .-.Technical and methodological matters with, respect to the formulation of a sys.tem of quantity and price index numbers, and
the corresponding collection and compilation of price and quantity series?
are .dea^t wi.th. in the.two papers prepared "by the ..United.Nations Statistical
Offioe0.l/." A paper-on a preliminary study of the. quantities, producer^
values and derived'unit.values of agricultural commodities for selected African countries as.available from, national accounts sources, has
also been prepared as a separate document0£/ ..
20. In this, paper,_country, practices have largely been delegated to an. Annex, in which, some basic/' characteristics of index numbers of prices and quanta available in African countries such as official title of index, scope and coverage, original weight base, revision of weights,
foiGiula used,, number of items (or commodities) included, .frequency of
compilation, "peripd" available, type of'weights used, -data; collection procedures., .are given in country and index detail.'
3° . As a result, the written text-will necessarily be brief. The paper starts with a review of the g-eneral availability-of index numbers of prices and quanta, in countries of the region, w.ith.the' object. of pin pointing the more obvious gaps in statistical,data? to.be-followed by a desciption of national practices in the collection of.data on prices and quantities and the corresponding compilation of index numbers in various fields, under the headings of producer prices, retaiT'or 'consumer prices,"production statistics, external trade indices, and output of producers - of government services»
■:4'* !.■:.;■■ The.".material, used. In .the. preparation of ■ the paper, has largely been
drawn: .from country replies to an ECA Questionnaire on Statistics of prices'. ,and: Quanta-, dated February 1969, for which a total of 32 returns have. been.received~. and.from ■.methodological descriptipris and notes .
1/ A Draft System of Quantity and Price Index Numbers (e/CJST,3/4O1) 5
The Collection and Compilation of Price and Quantity Series
IE7OST374O2).
2/ A Preliminary Study of Quantities and Unit Values of Agricultural
Commodities used in the Compilation of National Accounts.
Page
contained in United Nations statistical publications 1/ as well as in those issued by the countries^/available in the ECA'Library. As in the nature of things, the material and information available is
necessarily incomplete; and the ECA secretariat -will- be grateful to be .kept informed of "deficiencies and inaccuracies coritained in this paper .so, that later revision of "/it can be made. Ike /secretariat: should also
■like to express- its'gr&vj,lade jo the countries for their co-operation
in the ■ inquirye ■ ' ■ . ",'■'■ ■'.-■ ■ ■ ■" ■ ■■ ' ', " ■ '/, ;-. .v ' v v ■ ■ ■' ■
-5*-:y 'In-using '^he-paper, ■Reference"should also be' made ;to" 'the' following TJm-ted Nations Statistical Office or ECA documents, amongst o'tners:.
(a) Jata Sources:aM Country Practices in the Compilation of
; ■ ^^^-■S£§lrl^y^E_^d^elate:d Statistics in, Africa ■ (E/OT.14/MC/30?- October 1968); Annex I, for infoSnation" ' on over 100 household consumption and expenditure surveys,
held in Africa5 ■ ■ ' '
(b) / Consumer Price Indices in Africa (B/CN.14/CAS.5/LS.l6/Rey.-l,
August I967), for' country descriptions of consumer price ' - indices compiled in 39 African countries5
(c) ffuidelines for Annual-1 and More Frequent Industrial Inquiries
in African Countries f Tg/fTff. 1 /] /wiv/z . ^^^^ 1?tf) ^ '-j^pTr .- prepared by the United Nations Statistical Office, for Tables 1,2 and 3 which give information on country practices, in the - , ■ -collection of industrial statistics in Africa? and Index : :■■ '^^s °? Industrial Production (E/cij;i4/STAT/6 5 November
19-66), a papers prepared ^ ECA secretariat,- for selected .. ■, ■ information on indices of industrial production compiled in
the African region* ' ' •■''■:.
The.publications include, United Nations Statistical Yearbooks,- : m Monthly Bulletins' of Statistics9:ijS-Yearbooks of International
Trade Statistics; the I.967 Supplement to the Statistical Yearbook
and the Monthly Bulletin of Statistics: Methodology and Dsflrii +A nng
issued by the UK Statistical Office * the 19 66 Production Yearbook .
of the Food and Agriculture Organization, and Technical Guide's
Consumer.Prices, Vol. qt Descriptions of series publishedirTthe
Bulletin of Statistics..International Labour Officef 1968,
Page 3
ECA
6. . An BCA Questionnaire on Statistics of Prices and Quanta was sent
to African countries in February I969, and at.the.time of writing.a total of 32 returns have "been received, including 5 "nil" replies, that is, replies concerning countries for which, no- index numbers of importance have so far "been compiled. The distribution of African countries bysub-region with respect to the completion.of the EGA questionnaire is
given "below:
. ; .' ,' TABLE 1 ' ■ ■ ■ "
Distribution of African countries by. sub-region with respect ' ■ ■ .fo completion of the ECA questionnaire, . - .
.. North West East Central Other. Total
, ' Afri.oa Africa .Africa Africa Africa- '
Number of countries " . . .
that replied .. , . 4 13 8 5 2 ■' 32
tanber of countries . . '
that have not replied ... 2 1 1 3 3 10
dumber of countries not,
included in the inquiry - . 1 1 _ 1 3
15 10 . 8 6 45
± Eefers to Portuguese Guinea, Southern' Khodesia and the Eepublio of
■■ South Africa.
7. The-country replies received vary considerably in quality and.degree of completeness. On the whole, they do contain up-to-date information in the majority .of cases. Gaps in information through incomplete coverage of country returns have to a large extent, "been made good, by making use of national as well as international publications, a fact already referred to in paragraph 4. The names of countries -that replied to the -3JCA questionnaire can be seen from Table 2, while country programmes for the establishment of an integrated system of' index numbers'.of prices and quanta, where available, will -be summarised in a later section of
the paper. . ■ . "
8. The ECA questionnaire of February I969 requested, the types of data as listed in paragraph 2 with respect to eac-h index number compiled.
The types of indices covered include those of intermediate consumption . .
as well as categories of final demand (i.e. final consumption,, fixed . . ■
capital formation, exports), volume of production, producer and consumer '
E/CN-14/I9AC/34"
Page 4
wage and salary rates, and index numbers specially constructed ..for national accounts purposes. In addition, information on data
collection procedures, e.g. whether: sampling, methods are used in select ing outlets, and whether .the method of direct purchase and the method of. questioning purchasers' have been used in gathering consumer prices,' etc., has also been collected. ■ . .
AVAILABILITY'OF DATA;" '. ' ' .
9- As expected, the availability of data on prices and quanta in
African countri.es appears to have been highly correlated with the stages oi statistical development in which the countries find themselves. A bird .s~eye-v3.ew,;o.f index, numbers of prices; and-.quanta oompiled..,i&:'-
countries 'Of the region is provided \>j Table. :2*-' ■
,10... .In. general,, it. would seem-that African countries have made" fairly "' ..good prpgress.:: some 53 consumer price indices are being compiled, of
whxch:34.are for-African households and-19 for expatriate (or European) .households.|. .15 national■ wholesale price indices "are available;" 8'Voun-
tries have indices of agricultural production and 15 countries, have ' ' compiled indices of industrial production; while no less than 36 African countries possess index numbers of quantum and unit value of -imports :■
and exports.
11. Or. closer examination, however, the picture appears to be much'less satisfactory, In fact, great deficiencies in statistical data are"found' ..to exL.s.1^. g0. far.,as the-indez 'numbers of prices and quanta are concerned,
a more realistic assessment of the situation is as follows: ' '
(a) Although 53 national consumer price indices are compiled, only
•34 of them refer to the consumption pattern of African house holds. In addition to the 34,, six more countries, vis., Witania, Senegal, Chad, Congo (Brazzaville), Togo and bouth Africa,.have bo far compiled consumer price indices for European-households only, (or urban white population, in the case of South Africa).. Of the'40 African countries for ■ which such indices are available, be they for African or
'■" ?iro^fn h°£sehold35 only three countries (i.e., United Arab < =
■ Republic, Ghana and Mauritius) have compiled indices of .- national scope5 in the' sense >h&t both/urban and :rural areas.
ot_ the country are covered on a. .national scalej .4 morecoun- ■ , ries 1/ ^i.e., Nigeria, Southern Rhodesia,.Zambia and South ' Attfiea; have compiled indices for the main towns? while the remaining 33 countries have calculated indices for their respective capital cities only. The latter category of
indices includes some the., weighting system for-which has been __^_^ based °^ rather questionable foundations 5 ■■
as envisaged
^l b* ^f^^^is category if the intended consumer .
ged Including Dar-es-Salaam) becomes available
TABLE 2 / Page"5 Summary of Index.Numbed of Prices and Qianta oompiled
(in "a:" indicates that a particular index has teen compiled and available)
Sub-region and Country
Countries that returned ECA
questionnaire */
Wholeeal price indioee
Retail or consumer
.price indices
External trade 1^
indices Unit Value
Indices of Production
aH O 01 [Q 4)m oH tJ ol
<HO -da)
■rl PiB
fto
Pi n
§o o
(1) (2) (3) (4) (5) (6)
fforth Africa*
Algeria Libya Morocco
Sudan „... * Tunisia
UAH (Egypt) West,Africai
Dahomey Qamtda <
Ghana .«
Quinea Ivory Coast Liberia
Kali ...
Mauritania
Kiger -.. .*...
Kigeria
Portuguese Quinea Senegal
Sierra Leone■ ...-
. Togo' ,.'
Upper Totta ...1....
Cape Verde East Africai Ethiopia
Kenya .
Madagascar Malawi Mauritius Somalia
Southern Rhodesia Tanzania
Uganda Zambia Reunion Comoro Islands
Central Africai .., . ....
Burundi .'„..
Cameroon
Central African Republi Chad
Congo(Brazz. ) Congo (Kinshasa) Gabon
Rwanda Other Africai
Angola Botswana I«sotho Mozambique
South Africa (Rep. of)..
Hgwane (Swaziland) Totals t
(7) (8) (9) (10) (11) (12) (13) (14)
3/
rc/
XT/
rn/ xn/
**/
xm/
ly
xt/ xt/15 34 19 36 11 11
ft Excludes a return from tae East Afrioan Community, which compiles axternal trade indioes for Kenya, Uganda and Tanzania.
Page 6
Footnotes to Table 2
a/ EGA questionnaire of February 1969, on Index Humbers of Prices and
Qianta. :
b/ Includes both imports and exports of merchandise.
0/ Wholesale price index of imported consumer"goods.
d/ Wholesale price index of building materials.
e/ Wholesale price index of imports and exports. . f/ Wholesale price index of government stores.
gj (k) means national in soope.
h/ Calculated by France for these countries.
±/ The indices refer to exports only.
2] The indices refer to vexj early periods only, i.e. 1937-1947*
k/ The indices are calculated by FAO on a uniform basis employing regional
constant weights (Source: FAO Production Yearbook, 1966).
l/ Includes index numbers of the volume of gross agricultural production,
indices of the volume of net agricultural production, and indices of the value added at constant prices.m/ Production index of European agriculture.
n/ Available for 1949-1962, but index discontinued.
0/ Index numbers of the volume of inputs of agricultural production.
2/ Indices calculated for 1955-1963, but discontinued.
_q/ Cost index of building materials and labour.
r/ For imports only (for 1963-65). The indices have been constructed
specially for national accounts purposes.
s/ Indices of wages and earnings for employees on Sugar Millers' ' ' .
Estates. ' ■
\f Base year 1964. However, the indices do not seem to have "been
officially rel-eased.
E/CN.14/HAC/34
Page 7
(h) .Of the 1-5 national whole'sale price indices-listed in Column (2)
■ . of Table 2, one refers to imported and exported goods ("for Ethiopia)? goods produced for local consumption not entering
into .the calculation of the indices 5 one for government stores
: . (Malawi)5 another for imported consumer, goods (Liberia); and
three" others are confined to wholesale prices of building
materials only (Southern Bhodesia, Zambia and Senegal). So
properly speaking, only 9 -countries:in the region have compiled -■wholesale price indices covering both import^ and local pro-
; ■ duction;- ■ ■ '
(o) Of the 36 countries" for which external trade indices have been compiled, no less than 11 countries (mainly in the French-:
speaking group) had these' indices compiled by an external
source. In addition, these indices for 3 more countries have been discontinued. So effectively we have more or less regularexternal trade indices compiled by 22 African countries' only5
:(d) Of the 8 country indices on agricultural production (See Col. 7 . of Table 2), 5 sxe calculated by the United lations Food-and
Agriculture Organization. So effectively national indices -■are available for 3 countries only,'viz. ?: .United Arab Republic,
South Africa and Southern Rhodesia? with the index for the last—mentioned country referring to European agriculture
■ only; . ■ - " ;. ■ ■
(e) Of the 15 countries which possess indices on industrial pro
duction, 11 have calculated indices for mining.,1..15 for manu facturing, 11 for electricity and gas, and 2 only for construc- . tion. The list of countries'that have compiled these indices
; is shown in Columns '-8 through 11 of Table 2. • It can be seen
■ from .the table that these indices are centred in countries of the: Jorth African sub—region, :with a few notable additions .-. frora-the other sub—regions such as Ghana^ Nigeria and Senegal
. from ¥est."Af ricaf Ethiopia^ Kenya, Southern -Ehodesia and
.. ■ ..: Zambia from East Africa? Congo (Kinshasa) from Central Africa;
■ .. and Mozambique and the Republic of South Africa from Other
. " ■ Africa; " ' .-..:■..- . " ■ '■•.-■■
(f) Column (l2) of Table 2 indicates that the United-Arab Republic
'appears to be the only country that has compiled an annual index of intermediate consumption for the agriculture" sector
(i*e. volume of inputs -Used in agricultural production)-^
although it is known that several other countries (e.g., Ethiopia, Tanzania, etc.) have made rough estimates of the
same item for double—deflation purposes in order to arrive _at constant price estimates of value added by the agriculture
sector. 2fo indices of prices or volume of intermediate con sumption appear to have been compiled for the other sectors, although this may be a feasible thing to do for some African countries through their periodic compilation of input—output tables;
E/CN.14/NAC/34
Page 8
■■'■- (g) A few. countries in the: region- have'calculated indices of. wage
■ and salary rates (e.g., Kenya, Ghana, Mauritius'/ United Arab Republic)-.for one purpose or another. It is' known that ad hoc
. ■ :. : index numbers of wage, rates and earnings for selected kinds-■■"- ■ 01> activity (e.g., for employees in public■ administration,
; " ■■ construction, manufacturing, etc.) have "been compiled for
/.national accounts purposes "by a' number of'countries in the■;.■■■...■. region, including Ghana, Nigeria, Malawi, Southern Rhodesia,
-; ■■■' . ■ . Zambia, South Africa and Tanzania. , These indices, however,
"- ■ have not been compiled on a regular■ and peribdic'basis.
12• ■ The.:inadequacy"of basic data on prices and quanta'and related indices in-the region, would be better appreciated if the matter is■[looked into on a country-by-country basis. Thus, 'at least "six countries l/ in the
■region .have not so far calculated- index numbers and the list would be lengthened "by the inclusion of' seven, more .countries 2/ if.their consumer price indices, be they .for African or Shropean households, are left out
of account, , The current situation for some of the countries which have not yet calculated index numbers of prices and quanta, is summarized
below: .
■ (a) . Botswana:. A cost of living index is currently being prepared
and intended to be compiled on a quarterly basis for the country .as a whole. The weights and- selected-items for the index will
be derived from a household expenditure survey currently in
progress;
Dahomey; Ho indices of prices or volume have been compiled
as yet; but the necessary raw data on prices have been collected, which will enable' calculation of a number of indices to be
made, including a consumer price index for African households, a price index for building materials'and external trade indices.
Retail prices were collected at a large number of markets during
the 1965-66 national Household Budget -and Pood Consumption
■■Survey, which included-a random sample of 659 urban and 2,201 rural households. The- weighting problem for a consumer price index would be solved once the results of the I965-66 survey have become available. In addition, monthly, collection of prices of construction materials in Cotonou was started in February, I967"; , '
1/ pahomeyj' Gambiai ■■ Guinea, Botswana, Lesotho and Rwanda* ■
2/ Congo (Brazzaville), Chad, Mali, Mauritania, ITgwarie (Swaziland),
Portuguese" Guinea and Somalia.
Page 9
(■°) Guinea: Hie country_is now in a preparatory stage with respect
to.surveys of prices and quanta?" establishment of national accounts in sector detail for. the year I967 is known to be in progress;Lesothoi Collection of." elementary series has begun and calcula tion of some indices might "be undertaken in the, near future?
(e) Rwanda: No price indices or indices of volume have been
compiled so far due to lack of surveys undertaken. However, collection of prices at wholesale and retail levels and of building materials has been started.13, Relatively few countries in their replies to the EGA-questionnaire, referred.to their future plans or programmes with respect to the establish ment of an integrated system of prices and quanta within the framework of national accounts. Among those which did give a reply on this score, the answer was usually a negative one. However, the following cases
may be of interest and worth noting: ■ , .
(a) Algeria: Apart from the continuing compilation of indices
of wholesale and retail prices and indices of external trade, an index of industrial production is in the process of being elaborated and reconstructed;
(*>) Ghanat A scheme for the establishment of an integrated system
... of index numbers of prioes and quanta within the framework . of national accounts and for the improvement of existing . indices, is being worked out? ...
■ (°) Liberia* Hie country intends to develop index numbers of
. : wholesale prices of export goods and of import goods, and-
index1 numbers'of quantum and. value of iron ore and' of. rubber j (d) Tanzanias A consumer price index for the" urban areas of the
country on a national scale, consisting of 18 towns (including Dar—es—Salaam) is in process.of being calculated and is due
to appear in the course of I969. A total of 128 items will be priced for the index. Th.e collection of prices is made on a monthly basis for. such items, as are purchased from the Central Markets and on a quarterly basis for other items.
A system of weighting together the items and-towns, is expected to be provided from a pilot investigation of the "Household
Budget Survey of Tanzania, 1968/69"; ' '
Research into the. calculation of a consumer price index for African households .is currently in progress. (The weights and items for pricing will be derived from the results of. the
national' Household Budget and Consumption Survey, 1964-65,
which is of national coverage, and included a total, of 1,600- urban and rural households;Page 10
(f) United Ara]3_Republac_g A series of new indices has recently
been calculated, which necessitated-the collection of addi tional series of data* =. -Theae indices range from 'new ventures
■ in the field of calculating price and volume indices for the .agriculture sector ■to.substantial revisions of indices of'
wholesale, and■• re.tail prices and of industrial-production,
■ ■ ■■■• The main reatureS" 6f these indices are as follows: '
; ~ Indices for the agriculture sectors * These consist of ... - . ■ ..,,.5 indices;, -dealing respectively .with, the. volume of gross l
and nei production^ the volume of inputs^ and value added at "bcth current and constant prices, Double-.deflation (R.C -Geary's formula) has been.used.for the purposej '
■ - Consumer price indexs The new series of indices f 01 the urban and rural sectors respectively are of national scopes
■ . - "Wholesale price index* The new series is of .national ' -,
coverage and includes a classification "by end-usej
..- Index\of industrial production; The new-series uses quantities of production as indicators instead of the , ■: existing value added series' deflated by wholesale price
indices• ' ■ ■
■ Jlost of the new series of index numbers listed above have not.
■ .-.yet been published. Presumably a period of experimentation . , and consolidation'will be needed before the new system of
index numbers is finalized,
14- .There is little doubt that fresh developments of interest in the compilation of index numbers must have occurred recently in countries of the.region, These, howeverr cannot all be incorporated-'here on account of lack of space* Special features of their practices wi be dealt with under appropriate headings ir later Petitions''of the PRODUCER PRICES iTO PRICE IlTDICSSs " ■ *■■ " - " \'.
15. Although producer prices occupy a central place in the system of pnqe statistics and are useful for general Gconomic analyses and
studies of market conditions, suoh -statist:* os have s'o far been collected on a regular basis.in a limited number of African countries -only. Rather the emphasis has been on the collection of retail or consumer prices in connexion with the calculation of- consumer-price indices. As a matter of fact, even where producer's values of agricultural production have been included'in- the national.accounts statistics of a fairly large - number of .African ■countries, few of them'have defined explicitly the - basis of these estimates. In still a number of other African countries, no estimates of the producer's values of agricultural production have
yet been made* . . : '
Page" 11
16. -:So .;far..,as the. compilation of Index numbers of prices-in this field is:.concerned.): :they relate to -wholesale prices collected from importers, /wholesalers' and local manufacturers. As already mentioned .in paragraph
ll(b;):, pr.op.erly speaking, only 9 countries of. the .region have so far
compiled wholesale price indices covering "both imports and local pro
duction, l/ namely, Algeria, Central African Republic? Gabon,- Ghana,
Morocco., South Africa, Sudan, Tunisia, and the.United Arab- Republic Countrypractic.es in.the collection of. data and the. calculation of these index numbers are shown below? with respect to the scope- of index,
classification, basis'.of weighting., formula.used-, frequency of. compila tion,, .and data collections . . ■ ■ . , ' : • ; .
■- (a)- ■:■ Scope of index; : Geographically the indices .may-be limited
: in coverage to some capital cities'or relate to the whole country* For instance, the existing wholesale price index
.''■ :■' ': : :- for the TJAR (with 1939 = 100) 2/ reflects .changes in whole—
. . ,-■■■ sale prices in Cairo and Alexandriaj while the new index under
.-,:■."■; . - preparation (with 1965/66■■= 100) covers the.whole country. .
¥ith -the further exception ^of- South Africa,- indices for the remaining countries relate to their respective capital cities .
■■■..-■ ; - only (e.g., Accra for Ghana, Casablanca for Morocco, eto«)j (b) ■• Classification;. Two" approaches to the classification of
producer and related price indices are in use, vis., the traditional commodity approach and the sector approach* Ihe ,.i . ' <: former^is. based on a classification of commodities according
.to the composition and other physical; characteristics and . ■.their stages of. processing; while the latter approach is
. - . '.: - based: on a broad division of industrial activity (e.g«,
agriculture, manufacturing) in which the commodities are
produced and the economic origin and destination of the .' _ commodities.marketedo.r Classifications according to economic
origin generally distinguish "between domestic production and ■
.' .:.:-, importsi and classifications according to economic destinationusually distinguish between intermediate consumption, final . lf: .consumption^, fixed capital formation and exports. Among the
9 countries which compile wholesale price indices in the
. ■ ■ .region, all of them have adopted the'traditional commodity
approach to the system of class-ification. But a number of
1/ Excluding SJthiopiav the indices for which cover imported and exported
products onlye ,
g/ For the existing wholesale price index for the.UAE(with 19.39 = 100),
. the prices for cereals"are collected from wholesale outlets in all
governorates • ' . . '. ' ' ■
Page 12
countries have used "additional systems, of classification: ■■ ; .. for instance,. South..Africa-.has also used the ;sector classifica—
,. tion, .distinguishing a total of 6 sectors .for.the index,,viz., agriculture,' forostryP-fishing, mining and quarrying, manu—
. facturing, and electricity,and gaao A number of other
■ countries, falling short."of adopting the .sector:approach, have used a classification of commodities■ 'by economic origin
(i.e* distinguishing domestically produced from imported ■ ■ . ■ products.)-,. such as Gabon, Morocco and Tunisia. The ,UAR has
-.' ■ . also adopted" an end-uas classification- for'.its new-"wholesale ■price1 index, distinguishing 5 end-use categories, namely,
■ ■ Fuel, primary and-in":cr;::s:l\;-,';;o goixl.2, capital. goods, durable
; ;■ \; .consumer ,-godds, and non— durable, consumer.,: goods,- in-;.-addition /-.- to an 18—group commodity classification.for the index;
(q) Basis.■.of. weightings The bases of weighting used in compiling
. ' the index also reflects the latterrs scope of coverage, Roughly three types of weights have been used so far.in the 9 African, countries'' wholesale prices index calculations'!
(i) The .total supply of goods in the base ^year: Ghana
■and the UAR, for which the weights are based on the .... value of commodity supply for consumption and
. . exports (ioe° local production + imports);
(ii) The total.supply of goods for domestic use ■ (i.e.
. Local production ■■;-. imports — exports): The
. . countries using this concept include Algeria, Gabon, Morocco and Tunisia.; with their respective weighting basis as follows i
— Algeria? Value of total supply of,goods for
■ " domestic use (oxcluding subsistence consumption);
— Gabon? Value of -wholesale transactions (excluding exports)s
— Moroccos Value of goods domestically consumed;
. . — Tunisian The average annual value of■commodities marketed during 1936-1940, revalued at 1940
- prices. . , ■
(iii) Estimated quantities of goods sold in domestic'. ■ \ "
markets during 1953* Sudano .
(.d) Formula used: The Laspoyres formula, ±oe» based-weighted
arithmetic averages of price relatives, has been used through out, with one exception,. The existing wholesale price indox for the UAE represents a weighted geometric average of price relatives; however, in the revised version of the index, Laspeyres formula has bean usedg
E/CN.14/NAC/34 Page 13 '•''.. -
, , (e):, Frequency of compilation: ,With the exception of Gabon, which
■■ ',. ' compiles its index bi-monthly and Sudan,- which compiles the
index quarterly, the. remaining 7 countries,have ."been compil- ing the index on a monthly "basis;■■■'"■■ (f) Mta' collection: In therma;jority of cases, the data collected
"'■■ ■■■' ■ '" refer to producers< prices or wholesale prices for agricultural commodities; ex-factory prices for manufactured goods; and
•- ....wholesale traders' priGes for imported and exported goods.
';,.. . . However,J the.practices vary:from country to country. For
... : ...instance, quotations for imported and exported goods, in the .. i ..'/'*"-■ :.6ase.;of Algeria, Sudan.and Tunisia, are average o.i,f. and
' 'f.o.b. prices respectively from foreign trade .statistics.
Government-controlled prices were also used for some com-
■"' ' ; modities in'the case of Algeria.' Sampling'methods have not . "Been used in selecting price outlets. Rather, reporting
■ establishments were in most cases selected from among the :-.■ -•- "leading importers, wholesalers and manufacturers in the
;■■■■.> country. Specifications of commodities for pricing are . usually made in terms of physical characteristics, quality,
materials, model or type. In the.case of South Africa,
■:■-^--■■■ ■■ adjustments for seasonal fluctuations in the prices of fresh
■■■■■ ■ fruits: and vegetables are made by shifting the weights >of ' these seasonal'items- to other items in the relevant sub-group
■■' ■;'''^" -during- the off-season period, t&ile- the prices of seasonal commodities manufactured are assumed to-change in accordance with price trends in the related items during the period of ' :':' 1; ■■"■ ; seasonal disappearance.
SETAIL OR GOHSUMEH PRICES AID PRICE INDICES:
17, There are altogether 40 African countries which have been collect ing data more or less regularly on consumer prices and compiled consumer
price indices of one type or another (See columns 3 & 4 of Table 2 on
p.age- 5)-. Some ■ of these indices cover no. more than foodstuffs only(e.g.-Mali), or with an'additional group fpr fuel,... lighting and heating (e.g.,: Angola, Cape Verte, Portuguese Guinea) | while some others refer1
to expatriate (i.e. Euxopean) households only, with no corresponding indices for the indigenous or African population. The latter1"categoryof countries'1 include Chad, Congo (Brazzaville), Mauritania, Senegal, :-.
Tbgo""-a!hd the Republic of So'uth Africa,!/ Country practices in the
collection of data'"on consumer prices and the calculation--of price
indices in countries of the region will be- summarized below with' respect to such items as'the scope of the index, formula used-, • data oollection?
quotations on'rent, -seasonal adjustments'to prices of selected commodi ties, and frequency of compilation* ' ■
l/ The consumer price indices compiled by South Africa refer to the
consumption pattern of white population only.E/CN.14/HAC/34
Page 14
(a) Scope of index:—* Geographically "the .scope' of—consumer price
indices compiled in African countries may "belong to one of ' _ the following three categories: ■ ' '
- ■— Indices of national coverage, with separate.indices for
•'■• ~~ -. both urban and rural sectors: .e.g.',' Ghana? Mauritius, and the United Arab Republic; ■■• . . . ■ •■■..-,..
- Indices covering urban centres only! e.g. Nigeria (5 towns), ,; South Africa (9 principal urban areas); Southern Rhodesia .
;';:'.';"';,::: :(4:t(pwns)^ Tanzania (18 towns includijig.JDar-ee-Salaam, t * ■ for the intended index), and Zambia (8 towns';, 4 of which
are copper-mining ones) 5 ' .,,•' ■ - . ■
..- Indices covering the capital-city.only. ;To this category .. ■ almost all the remaining, countries1 indices belong.
* ■ .
So far as the sections of the population covered are concerned,
the consumer price indices- of African countries may be classifiedinto the following categories: ■ .
(i) Indices'covering .all indigenous households: Included
in- this category are the indices for 10 African ' countries, namely, Algeria, Cameroon,■ Ethiopia,
-; Ghana, Mali, Madagascar, Mauritius, Niger^ Southern
■. - ' Rhodesia and the UAR; .
(ii) Indices referring to households of■wage—earners and
salaried employees or low-income households: Included
in this category are 15 African'countriesr indices, via*, Gabon, Kenya, Ivory Coast, Liberia, Libya, - ' Malawi, Morocco, Mgeria, Ugwane (Swaziland).,.Sierra Leone, Tunisia, Tanzania,. Uganda, Upper. ■
■ ■ Vo-lta .and Zambia; ■ ■..._:
(iii) Indices referring to middle-income (or middle—class)
households.: Included in this are Angola, Central ■'
African. Republic and Somalia; . ■- .
Xj. /fflaerever more than; oaie-.consumer price indices are compiled in a country, the descriptions under- this heading refer:to the principal index only, i.e* they refer to the index for African households - wherever separate; indices-' for both African and European households'.
. ■ are. available. -. In case ^separate indices for the .lower-income- and
middle-income groups are compiled in a country, the descriptions
refer to the index for the lower-income group, etc.\ ■ ■ , . ■E/Cff.l4/NAC/34
Page "15
.Indices referring to -government employees: e.g*
the indices for Sudan, Burundi and the Democratic
;__.;;:. ■; . i;i -.-Republic. o.f -t Congo; ■■. .. -r -.;■,:■■.... ■■ ■ .-.-■". ' '''■
_...("v).;Indices for expatriate (European) households only;
■ .., ; ,...■■ Included in...this: category -are 6 countries' indices,
. _■_._■ ; - .. . jit±z».y -Ghadj ;Gongo (Srazaaville)9 Mauritania?
... - ;. ,. Senegal j Togo and South .Africa; ; ■■ -,..■■
-■.■ ,t .(yi) . Indices the coverage .-of which with-respect to
- .: ,. ,.,..■,. .sections of the population is unknown: ..;..-Qape
-..-,.- Verte? Mozamhitiue and Portuguese Guinea^.'-.
formula u@.&d: J'fith one known exoeption?l/ the,.Laspeyre formula (base-weigtxted arithme-tic averages of: price relatives) has
been ;us,ed by :all. the ,-Afrio.an-countries in .-compiling their
■-consumer -pripe indices. However, .variants of^Laspeyre's formula have-"been;adopted;-in the .oases of. .Libya and,Sierra Leone ,as
";- (l) : ':li±%yki'■■■ Ghaihed nretHod ,is used in calculating price
■ ■'■•'relatives f or: all items, the formula-used being:
-"■-■ < Where i(p') stands for the consumer'price index;
'" --P'Qr.t'"'ar-e -the-prices and quah'titles" in" the base year .0;
r: ■ ... ■.!> s Pn_-i a.re . "fcne Prices .for periods n. and1 n— 1
■'■■ 'respectively, " '■ ■ ■ ■ ■'■.■- ' ' ■ ■ ■-
(ii') Sierra Leone: Chained-method is also used: in the
■ ■ :- case1.of Sierra Leone, with ."the difference "that
■ : inst'ead.1 of using the values of' .oonsumptipn expend!—
• . '. ■ ■-. ture on items in the base period' (i.-e./.p : o ) as
weights, a mixture has been introduced as follows:/p
n-1 ' P..1/ The- consumer'.price^index for Cape V&cte (Praia), represents a-simple
■ unweighted arithmetic average of price relatives., . .. -; .-.■•. ■■■; ■■
E/CUT.14/NAC/34
Page;16 . . , ;■..;■
Where the symbols l(P), p , q, , p - all
- ■-■,■ .■.■:= have.'■ the ! same meanings as in the cas{3 of Libya.
(c) Data collections Sampling;, methods,"based on probability have
not ~been applied in most cases for the selection of retail
*■ •-..= -price' ■'outlets;- ■ :The frequency of-price.'collection, i.e,,
■...-. daily, weekly, ^mo.nthly,; quarterly.! or annually, has been -■ .determined.largely .on the basis .of the degree of price
variability of .the. items to be priced. .In most cases, food items are priced daily or weekly^ clothing or.other items --, ^monthly; and. rent quotations-, rates and domestic servants*
. ^ages quarterly,, semi-^annually or annually depending on-the individual items priced and the source of data. The selection of items is usually derived from the relevant household
■ . expend!tur.e .surveys.conducted. In the -ca/Sje .of. Souih Africa,
■ ;. "' . all the items accounting for:'0.05$ or above. - of ■ th@ total
: . , . consumption-..,expenditure were ' seledted for pricings Information, on retail prices is obtained from major markets,
. > :..re.tail:-.stores- as well as service .establishments.1' A- combina tion of personal interview and mail questionnaire "is used . ,in. the majority of cases. Direct purchases by agents are ' reported to have been practised, by. 8...cptin:tries? viz., Congo
(Kinshasa), Ethiopia, Malawi, Madagascar, Southern Ehodesia,
Nigeria, Tanzania, and "Upper Vplta; although many more
countries must have be'en employing this'method. Often this
proves to be the only'way of obtaining accurate and reliableinformation on prices for .selected commodities on sale in the markets, from which the bulk of African households make
,. ... their daily purchases.. • The method' of,, questioning purchasers
■^ "is, also repor..te-d. to have... been, used, e.g..,. in Burundi.
Apparently, this method has not "been much favoured by coun-
• -.tries of the region. In those cases- in" which the chain method is used, as in Libya and Sierra; Leone,the.introduction of new items has given rise to little difficulty. For Libya, quality specifications of many items.are left:to the pricing .agents... The type and qualities for'which, prices are collected
^ are those'available in the shops, provided that some of these should be" the/same.as for the.preceding month to enable
calculations of meaningful price relatives,
Quotations on Rents The collection of quotations on rent has presented special-problems,for most African countries.
The difficulty appears to lie in obtaining comparable data for the same quality of housing over time, and in obtaining the right type of information reflecting changes in the cost of housing relevant to- the special consumer group for'wftioE ' the- indices, are compiled. Thus, 16 Ifficah countries. 1/ have'-
1/ Angola, Central African Republic, Congo (Brazzaville), Congo(Kinshasa)
Chad, Gabon, Kenya,, Mali, Mauritania, Madagascar, STgwane (Swaziland), •
Niger, Senegal, Tanzania, Uganda and Zambia.
Page 17
excluded it as an index item, nor has it "been included in
the determination of weights. Two more, countries,.(Ethiopia and Somalia) have included rent as an index item, "but have so far excluded it from the index calculations "because the
collection of 'rent quotations ,has no.t yet .started. Of the countries that collected regularly information on rent, national'practices for a total of 12' countries,'.are available,They are summarized into the following groupings:
■ (i) Eent quotations collected quarterly: Libya; Sierra
Leone, Sudan and Tunisia. Eent,quotations are
collected from 91 dwellings for"Freetown'(Sierra Leone), and from a'sample of 48 houses for Tripoli
Town (Libya); .,..-.
(it) Rent quotations collected; twice a year:-Liberia;
(iii)■■ Eent quotations collected annually:..Mauritius and
■' ■ South Africa. The quotations"are collected annully
■from I5G dwellings included in the l£6l-62 family
"budget survey, of Mauritius;
(iv) Eent quotations obtained from government laws and
regulations: UAR$
(v) Rent quotations referring to specified type and
1 size of housing 3-room apartments with simple
comfort for Algiers (Algeria), and moderate-rented
■ apartments for Rabat (Morocco).
The data collection procedures for rent quotations for ':2ripoli Town (Libya) deserve some mention. Two sources of
data are used for the compilation of' the ^rent index: one is based on a quota sample of 244 houses in .'Tripoli Town
■selected from the frame provided by' the I964 Census of buildings and Dwellings,' subject to two oontrois, viz,?
administrative division and size of family. Householders
in the selected, sample' were requested to furnish data on
monthly rents paid for January.to August, 1964, for the houses occupied by them at the time'of the Census. Promthese returns, the average rent per room was calculated for
each -quarter. As from .the last quarter of I964, a sub-sample of 48 houses was selected from the original sample.
This sub-sample is utilized to trace, changes in average
rent per room at quarterly intervals, with the first quarter
of 1964 taken as base.. The rent index is now constructed
by using the simple.arithmetic average1 of rent relatives for
each quarter by the chain method. A second rent index relates
to changes in newly, rented houses,. based on rents paid bythe State Property department for:newly rented houses each
month.. This index, however, was found to be not accurateE/CU.14/STAC/34
Page 18
■ " enough;1 da 'account of the changes, from one. 'quarter to another in. the housing standards of the houses reported in the quarterly rlists> From-January. I966 on, therefore, a monthly systematic
sample of houses' selected from :the monthly lists prepared by the Electricity Corporation, has been used instead as
source of data. "Such lists■contain, names and addresses of householders who apply for the.'installation of new electric meters. Ihe majority of these houses are newly rented.ones,the tenants of which, were requested to provide data on rents paid by them. A' quarterly index is then, constructed, again '■ .'■ usingthe chain method* . : !...
(g) Adjustments for seasonality of items: Seasonal fluctuations
. ; ;.. in the prices of fresh fruits and vegetables .and certain
■' clothing items during their respective'off-season periods often need to be adjusted before the computation of the related^price indices is made.. National, practices in this
respect for 15 African countries are summarized below:(i) Seasonal fluctuations in prices are accounted for
by omitting the seasonal item during its off-season
period and assigning its weight to the remaining. items within the constant group weights, i.e., by
shifting monthly item weights within the constant group weights, and/or by computing moving averages of the unadjusted indices: This covers the practices of 11 countries out of a total of 15 countries for which information in this respect is available.
The countries included are: Ghana; Kenya (12-month . moving averages for all vegetables and charcoal,
and 3-month moving averages,for. certain clothing
: - items); Libya (price relatives for summer clothes for May to October, those for winter clothes for November to April, and other clothing items for the whole year)? Madagascar5.Morocco (3-month . , -moving averages) 1 Sierra Leone;. Southern Ehodesia
(adjustments made for the index for Europeans only)5 South Africa (During the off-season period, the '. weights assigned to seasonal items are distributed'
proportionally among other, items, in the' same sub-
■ groups or groups); Tunisia; Uganda.(12-month moving
"<■'.' ■' averages); and the tfAR (for its new consumer price '
indices only).
: (ii) Prices are collected, monthly according to a mobile
basket of fresh"fruits and vegetables: Algeria;
(iii) Seasonal price fluctuations.are accounted for by
maintaining during the off-season period the closing
. prices of the preceding season: Zambia;
E/CN.14/NAC/34
Page 19
" (iv) Seasonal fluctuations in prices of fresh' Vegetables ' .'; ■■■'■• 'are accounted'for by using a sliding scale of Weights
■ ■■■ ■ ' ■ 'Mauritius; 'and ■■ _ " ' . ' ' ' (v) The choice of items in the' vegetable group-has-been
- ■ made so as to^ ensure uniformity all the year -rouiidt
"' ■' ' ' Congo (Brazzaville). ■ ■ .■'..-
(h) Frequency of compilation;■ Apart from one. consumer price -index,which is compiled on a bi-monthly "basis, i.e., the index for
~ . "the rural population of .the UAR,1 and 5 other -indices, which- ; ' ., are compiled on a quarterly "basis, all the remaining consumer
/ price indices, for African countries have been computed on .,.' ■ a,monthly basis. The indices compiled quarterly include; the
■' following: / .-.:-. ..--..
"■- Burundi (Buoumbura; Africans )| ■ ■ ■ ; : ' , ■' - Sierra Leone" (Freetown).; . , .!T . : ■ : ;.
-. Sothern Ehodesia (the index for Africans);
■:;,-.. -. .- Swaziland (Mbabane -&. .Manzizd} j~-and-". " " .'■""""' Tanzania (Dar-es—Salaam,1 for middle—grade civil s'ervants) * ■ ■'.■■■ '' .' .
P.ROSUCTIOIT STATISTICS ATO INDICES OF PRODUCTIOIT:
18, 'Statistics on production collected by countries of the region
include the gross output of goods and services produced, and the value a,dded, ~by "industries" whose production is intended primarily to cover cJds'ts:'of production. These include the activities of .the goods^-
p^odiicing industries, the services of distribution and transport and
the services produced by other industries such as financial, insurance, real estate and business services, medical? dental and other healthservice's, recreational services, and personal and household services, 19'. Statistics available on production for African countries, as is bommon with most developing countries of the world, have' largely been confined to the goods—producing industries and selected service
industries such as transport and distribution. ■' Although a number of
African, countries have held sample censuses of agriculture, XJ firm l/ 21 African countries are known to have held so far agricultural
sample .censuses in the 196O's. They include; Algeria, Central
■African Republic,\Chad, Gabon, Kenya, Lesotho, Madagascar, Mali .Morocco,■ Libya^ ¥igerr Senegal,'Sierra Leone, Southern Hhodesia,
Sou^h Africa, 'Sudan, Swaziland, Togo, Tunisia, Uganda, and Upper
■Volta. '";: "■■'■■■■
E/GM.14/HAC/34
Page 20
■agricultural production figures have not been established. With respect to industrial production, .information on ;the related surveys and periodic inquiries held in African countries is contained in document E/C3SM4/SJEA.'r/5, a document already referred to in paragraph 5('o)« ' Statistics on trans
port .(rail, air and. sea, with ,the exception of road transport) are usually
:available, while countries ''of the region are poor in statistics of
distribution. Although 'trade margins have been estimated for most
African countries for.national accounts purposes, few of these estimates
■have, been establlshed,"on. systematic, or .firm' foundations. . ' .
20'i;::;Sojfar::as.;p-roduction.;indices are concerned, ■■'only:a: total of 15 African
countries have so far compiled indices of industrial'production, and
only 3 of them, have calculated indices of agricultural production as
well-.- Hationai practices in the calculation of industrial as well as
agricultural production indices will be described separately with respect to such aspects as the scope of the index, formula used, indicators
used, systems of weighting and frequency of compilation.21. IHEEX NUMBERS OF iraUSTRIAL PRODUCTION:
(a) Scope of index; In the case of production indices for mining
. and electricity, the coverage of establishments is usuallycomplete on account of the small number of establishments involved. For the manufacturing industry, however, the indices cover establishments above a certain size only.
Handicraft establishments are normally excluded from the index. Thus, the following coverage 'by indices of production for the manufacturing industry has been reported for selected
■ African countries! ...
(i) Establishments employing 3 or more persons Or using power equipment or.steam boilers: South Afrioaj
. (ii) -Establishments employing 6 or more persons or using power equipment or steam boilers: Zambia;
(iii) Establishments employing 10 or more persons* Ethiopia, Morocco,' Southern Rhodesia and the United Arab Republic (iv) . Establishments employing 20 or more.persons: Libya;
(v) Establishments employing 30 or more persons: Ghana;
and . . ; _ . '"■ "
' . .,' .. (vi.) Establishments employing 50 or more -personss .■ ,"' '; .'. ' 'for which! the annual manufacturing .survey since
I963 has covered establishments employing'50 or
more persons only. ': . :
The geographical scope of the indices normally covers the whole country;
Page 21
(b) formula used: Laspeyres formula (base-weighted arithmetic averages of quantity relatives) has "been used throughout
■■■'.■ -; in the industrial production indices for the 15 African
■ ■ . .countries, which have compiled-these indices. Tnere are two 'exceptions, however, as follows:
■ ■■ -^ In the case of the UAH, the Paasche formula (current-
■. weighted-arithmetic averages) has been used in calculat—:
■ing 'the1 existing'-series of the'index1 of industrial pro—
■ .'■" duction.. For this, valu-e added-data are obtained "by agents-'from industrial establishments every 6 months, and the value added figures thus obtained.; are then
■1;':' ■ deflated by th.e wholesale price-indices to arrive at -■indices of produotion for the various industrial groups.
Laspeyres formula, however, has been used in calculating
, : . the new series of uAE's index of industrial production,
■. ." using value added data for the "base year as weights;
— Tkie indices, .of .physical, volume of manufacturing and mining production respectively for the Republic of South 1 Africa, have:employed -a,modification of the Laspeyres
formula* For the computation of these indices, the weighting is revised at the beginning of each year and is derived from .the results of the. Annual Industrial
■Census'for each preceding calendar year. The index is
■ . . calculated based on the fixed weights for .the preceding year, ■ .- " . "' -
(c) Indicators used: 9 African, countries have more or less
_;.' .specifically, mentioned the .type of series'used as indicators
. ' "in'the calculation, qf .-their indices of; industrial production^
The majority of them, have used almost exclusively the output
. (quantities) series, supplemented to a varying extent by other types of series such as-export/import series, input series
and employment,, series. . In-the case of-the existing index for the .United Arab Republic, and the index of gross output
in I962 prices.for Ghana, the indicators used consist
exclusively of value series (gross values of production for . Ghana and value added for the TJAS) deflated by their respective
wholesale price indices. Country practices in this respect are listed belows(i).. Congo:(Kinshasa):- Apparently only output series
^ ' have been.used;-- ■ ..._ - ■•
vi:0 Ethiopia: 35 output series were used as...indicators;
(iii) Ghana: For the monthly index for mining and
electricity, output (quantities' produced) series
only are used. For the' compilation .of. the quarterly , and. annual index of-gross' output in 1962 prices,Page 22
value series deflated by appropriate, price indices . are used. For ^mining and electricity1/ .the value - -.-...,■ .' series are deflated "by price .deflators obtained
on. the. basis, of ex-works prices of mining products and of electric enevgy sent out. For manufactur- i i - , ing as a whole, the deflators rused for the 4 groups
, of products, viz., beverages ancl tobacco, chemicals and vegetable oils, machinery'and-transport equip ment,, and other manufactured articles, are
- .. estimated on the basis of published series of the wholesale price index for Ghana;
■■;.-'■ ■-. ' (iv). .-Kenya::,- .For the computation,-of ...the-present "Quantity . . . :. . ■ thdex,.of Manufacturing Production'1, which replaces
■ . ■: . ■ " ,an earlier .index, the "Kenya Index of Manufacturing
Production'-'', 97-"volume of"output" series have been
■■;"■■ used as indicators. However, .of the. 97 series, / 2 series relating tp- electric.and other machinery
■; and 1 for miscellaneous manufacturing industries -- : - are proxy series, such'as consumption series or
a series of gr'oss domestic product in the group,
deflated by an index of average wages;(v) I'i'bya-i Only output series are used;
(vi) Nigeriaa Output series are,supplemented by export/
import series, while input (materials) and employ
ment series have been used sparingly. Imports. series have been used in the wood, paper- -"and printing;
■ rubber and chemicals, metal, and vehicle assembly
, industries; input series are mainly used for thefood, rubber and chemicals and the non-metallic
■minerals industries? while employment series are
.. used for two industrial groups only, namely, wood, - paper and-printing, and miscellaneous manufactures.■ An analysis of the types of series used by Nigeria
■in compiling its'Index of Manufacturing Production
is given belows
. ... TABLE 3
Nigeria: .Type of series used in calculating the Index of Manufacturing Production ;
O^rpe of series Ifaber of series Base year '- : . . .- .. ■■■ .-■ used . .'-■ weight(^)
Output series , ' \y 71.9
Input, series ... o... j:..'. a a
Export/import series .,.,..., ' , 1L -. 20.8
Employment series /g" "" ■ 2.9
" ' Total: " ' 38 100!o "
E/CN.14/KA.C/34
Page 23
' South Africa: 'For .mining, the indicators used are 'the quantities of.output for 60 mineral products '" ■"■ from all mining establishments* ■ For manufacturing, . ■ " '""'.' the indicators include quantities produced and
ancillary quantities of main materials consumed,
; the number of man-hours worked,, and the deflated
■'.'". gross values of production. 'For the pharmaceutical '•'■"'■ . Indus'try, the deflated value of sales is used;
(ri±%) ' Southern Rhodesia: The indicators used are monthly
■ ' ' '"■'■■" data'on output for 8 minerals and 230 manufactures.,
■; : and the number of'kilowatt-Hours of electricity
sent out. The data includes quantities produced,■"■■■■ ■' deflated value of - production-, deflated value of
■: '" "'■ ' ■ ' sales, quantities of materials used and number of
-'■■■ ■ man- or machine-hours worked; ."' (ix) United Arab Republic: The indicators used differ with respect to the existing and the new index .-• .' . of industrial production. For the. existing index,.
■ . . value, added series obtained.through half-yearly
inquiries, deflated by wholesale .price-.-indices, ■•■■
■■• ■ ■ " are the indicators. For the new index, the . indicators are the..output (quantities.) series-,^
'■'., -weightedby -value added-in the base year^.ox by ""~~ .
prires .in 'the base year,
(d) Systems of weighting: Three types-of weights have "been used
- in the'-computation or the 14 countries1., indices of industrial. production for which information is available.!/ As-io.which ■
- ■■■ of the three types of weights or a combination of them at different stages of the. aggregation,-is actually used by a
■ particular country is largely a matter of convenience and depends on data availability. The. types of weighting systems'
- used are as follows: ■ ' ■
(i) Gross valueo of production,(or gross output):
Ihe countries using it include Algeria, Ghana-■a&dL...
South Africaj "-.':■ .
(ii) Value added (or ne-t - output): The countries-using : '..this approach include Congo (Kinshasa) y Kenya?
1 Libya. Morocco, Nigeria,, Senegal, Southern Rhodesia?
. ■ . ■ : Tunisia, UJLR, and Zambiaf-and
(iii) Prices in base year: Ethiopia, The prices, are
, . : unit values derived t?y dividing quantities produced into the ex—works sales, values. . , Howevers. several
if Methodological information concerning the ±nd.ex of' industrial, pro
duction for Mozambique is not available at the time of writing.
Page 24
... , countries in the region have made use o:f. more than .... . . one., system' of weighting.- -.'^T^.nstance, although
■ . _., . .. Algeria, Ghana arid. jSbuth Africa have used a system of-weighting based i'o.ii gross values of production, all of them used value added as weights for1combining group indices into those for major groups? or for
> : .. combining indices^by kind of economic activity into a general index. The same may be said for Libya,
■ ■ . which assigned weights to individual industry groups
■ . ■ [ . ■ ■ based on the values of net output (i.e..,. value added),
,. .,. : ■ . . ; -while the internal weights %itnln; *each- industry group were; ■ assigned on the basis of values of gross output.
(e) ■ frequency ..of ..compilation; Of the '%% A£rican.countries which
' have compiled indices of industrial"production, 5 compiled their indices monthly,.4 quarterly and another 6 annually.
The availability of data has apparently put a serious restraint
on the frequency of compilation for these indices,22... IHDEX BJMBKRS OF AGEIOULTDEAL PRODUCTI-GH: : Country practices in the compilation of this index refer, to South Africa, Southern Rhodesia <
and the UAR only* - ■ . ■ . .
(a)5 Scope of indexs All indices .are. of 'national ooverage, except
the one for Southern Rhodesia, for which the ,index covers the production of, European agriculture, only; ;.
■■'W Formula usedi Laspeyres formula has been, used in;.-all. three
cases,, with the;average prices, for "individual "products
■ • ... ■ received by farmers as .-weights. ■ For the UARr an additional index-number for -value added at cons-fcant prices, has also
been calculated,. using Geary's formula.^. The- weights are provided by average prices received by .farmers in the case .of outputs, and average prices paid by farmers .in.the case
of inputs; ■■..-, . - ■
(°)' Frequency of compilation: Annual in..all three cases?
(3-) Jata collection: Quantities- of: output are gathered-annually
in all cases.
ESTERIAI. TRAD^! AID-RELATED QUANTUM AID MT VALUS JHDIOES:
23, ¥ith a few exceptions\lj almost all-.African countries have now compiled periodic external trade statistics. The coverage of data
1/ Countries which share'-common Customs frontiers &o not always have
separate external trade statistics in great commodity and country
of consignment detail, e.g.,- Botswana, Lesotho, Namibia (South-West
Afri-oa) and Swaziland, "■ ■-.. ' ' - .■■';■';, ■>■:■.>
Page 25
varies" from country to country; so-do the practices.in -valuing imports and exports. It can,.' however, "be said that,all countries, in the region are now' compiling-, their external trade statistics in accordance with
the commodity classification of the- SITC (Revised) l/ or with that of the1 BOS (Brussels -Tariff .Nomenclature).
24•- As already referred to in paragraph ll(c), index numbers-of■ external
trade-have "been compiled for .36 Afrioan countries, of which effectively only 22 are compiling their own indices of quantum and unit value on a regular "basis. National practices in this field -will "be described"below in respect of.such aspects as the.scope of the index (selection of items), classification, formula .used,'adjustment's for incomplete
coverage, introduction of" new items.,: and. frequency of compilation:■ fa) 'Scope of .index (.selection-of items);. The indices1 are normally
■" of' national coverage, with Customs declarations.of imports and exports as the source of material. Two criteria have apparently "been used "by African countries for the selection of import and".export, items for inclusion in the computations.
1 ' : These are: - " ■ ■; ■ ;'■■.;■■■ •"• ■
(.i) the inclusion of items, the import or export of
. -which exceeds a pre—determined level, in terms of
; ■ ■ . ' value: This can "be illustrated from, the examples
■ 'of the United Arab Republic and the member countries ' .'■ . of .the-East .African. Community, i.e., Kenya, Uganda
.' . . and Tanzania* For the UAH, all" items mth an annual export'value of L,E. 50,000.and above are 1 ■■■■■■ included in calculating the export indices; and
■ : , all items with an annual import value of L.E. 25*000 '■ ■ ' and above are included in calculating the import
. indices. In the case of Kenya, Uganda and Tanzania, only commodities with an export value of more than
£ 200,000 in each of the 3 countries and more than
> ^ "■ ■ ., . ■ £500..,000 in East Africa in any of the years I964-.
I967 have been directly used in the calculations
; L •■■■■■ . of indices for domestic exports; while for the ..■"■'. calculation< of indices for net imports, all items
with a total East African import value of less 1 ; . than £5°?000 in each of the years I964 - 67 have
been excluded. The most heterogeneous items have been rejected; :
l/ United Nations Standard International Trade Classification (Revised).
E/C3STU4/NAC/34
Page 26
.■.■";'■ - (ii) 'the, inclusion of only those. i1;em&" which aepount
for a minimum-percentage of the total value. , .
■•■'' -■ ■ of exports or imports^ .For Nigeria,., commodities . : ;. ■ .accounting-'fbr 1 per. cent or more of the total
value of exports ~or -imports in-1960, have ;been
. - . included in the index calculations. For SITC■Sections 6 and .7 imports, items accounting , for 0,5 per cent or more of .the total value "■
of imports have also "been, included.. ■ -
Classification.: A number of countries in. the. region .have
; '. ./"'calculated'-indices, of quantum and unit :value:;bf- External1 ■ '■
- ■ trade in accordance with the commodity classification of the SITC.. For South Africa, a classification; of external trade
■ ':"' " indices1 by industrial origin.) into, ag37icul.turalj;"miniing.-.and
■ quarrying, and manufacturing products, has beenused in
addition to a- classification into 8 groups."based: on the SITCj
(c) Formula used; ■ Of the 36 African countries, for which external
trade indices have been compiled either by a national institu tion or by an external source, 17 used the Laspeyre formula
(based—weighted arithmetic average of quantity relatives)
-.-■ for the quantum index and the Paasche- formula (current—weighted . ■ arithmetic average of unit value relatives) for the unit
■ . value index,, In most cases, the two index series are inter dependent. 9 other countries constituted exceptions to this
rule? while the indices for 12 more countries (including
11 African countries, the external trade indices for which.,
■' ■ ■ were calculated by France) used variants of the Pa&sche formula
>' . for the calculation of unit value indices. The practices . for the 9 countries which constituted exceptions to the rule,
are given below:
(i) Laspeyrefs formula for unit value, and Paasche
formula for quantum,.indices! Libya, Sudan;
(ii) Fisher Ideal formula for- both quantum and unit
value indices; Kenya, Mauritius, Southern Rhodesia, Tanzania and Uganda;
. (ill) Laspeyre !s formula for both quantum and unit
value indices: Angola, Mozambique.
The external trade indices for South Africa, so far as the index formulae used are concerned, represent a slight'departure from the' accepted rule, i.e.,:Laspeyre formula for quantum ;\
indices and Paasche formula for unit value indices. Beginning with 19^5? the quantum indices for that country are chained indices using fixed weights of the previous year, and the unit value indices are chained indices using current weights, A further departure from the rule is seen in the external