516 g4 to compare the INTRODUCTION constructed in ue Though originally index was level 3 , index ni rs Hallam pete index in 1750, BE PP ange, PACES cal devices and developed for measuring the em P'idely used "Newspapers, headline have today become Pa Ae they are fe Nae industrial prouucegy 1S the Tact that prices are going up oF GOST: “NE: reqsing, that crimes are ous perio rising or falling, that a to the beige che. econcrywerd by titer umber, They axe used to feel the Prnationary or deflationary Hig ie oom persed as indicators of inflatot rT omic activity’, ; ters: : ay .d as ‘baromel ye to an economy, fee ay eed See “fo what is happening to y {he index number of industrial he should look to important indices like the Tt otc, Production, agricultural production. aie are given below: Some prominent definitions of index numbers the magnitude of} 1. "Index numbers are devices for measuring eifferences ee a Cae en index nuraber fa statistical measure designed to show changes ina it ct to time, geographic loca- variable or a group of related variables with respe - eel tion or other characteristics such as income, profession, etc." Peg 3. "In its simplest form an index number is the ratio of two index numbers expressed as a per cent. An index number is a statistical measure—a measure designed to show changes in one variable or in a group of related variables over time, or with respect to geographic location, or in terms of some other| characteristics." —Patterson 4. "In its simplest form, an index number is nothing more than a relative number, or a ‘relative’ which expresses the relationship between two figures, where one of the figures is used as a base." —Morris Hamburg 5. “Generally speaking, index numbers meesure the size or magnitude of Some object at a particular point in time as a percentage of some base or| reference object in the past." —Herensontanenns It is clear from the above definition: s that an index number is ialize average designed to measure the change in a group of related vereneeed a Period of time. Thus when we say that the index number of wholesale prices is 112 for January 2008 compared to January 2007, it Means thi eis increase in the prices of wholesale commoditice 1. oeeest in the S to the extent of 12 per cent isa Beicy explained in the chapter on the iteme™ TPresenting a group of comparable; for nN of a certain locality — ha sa et nm se y va FL e be te ol seine ali ENi eX NUMBERS INDE e as no meaning at all. Fu poet tie hakeoe THOstarn eee teres ae must be the So eatictrsearibawr expressed in kg., Ib., ete. . g. lowever, this is not so with index numbers. Index numbers are used for purposes of comparison in situations where two or more series are cxpressed in different units or the series are composed of different types of ltems. For example, while constructing a consumer price index the various items are divided into broad heads, namely (i Food, (ii) Clothing, (iii) Fuel and Lighting, (iz) House Rent, and (v) Miscellaneous. These items are expressed in different units : thus, under the head ‘food’ wheat and rice may be quoted per quintal, ghee per kg., etc. Similarly, cloth may be measured in terms of metres. An average of all these items expressed in different units is obtained by using the technique of index numbers. (i) Index numbers measure the net change in a group of related variables. Since index numbers are essentially averages, they describe in one single figure the increase or decrease in a group of related variables under study. ‘The group of variables may be the prices of a specified set of commodities, the volume of production in different sectors, etc. Thus, if the consumer price index of working class for Delhi has gone up to 113 in February 2008 compared to February 2007, it means that there is a net increase of 13 per cent in the prices of commodities included in the index. Similarly, if the index of industrial production is 118 in 2008 compared to 2007 it means that there is a net increase in industrial production to the extent of 18 per cent. It should be carefully noted that even where an index is showing a net increase, it may include some items which have actually decreased in value and others which have remained constant. (ii) Index numbers measure the effect of changes over a period of time. Index numbers are most widely used for measuring changes over a period of time. ‘Thus we can find out the net change in agricultural prices from the beginning of First Plan period to the end of the Ninth Plan period, ie., from 1951 to 2001. Similarly, we can compare the agricultural production, industrial production, imports, exports, wages, etc., at two different times. However, it should be noted that index numbers not only measure changes over a period of time but also compare economic conditions of different locations, different industries, different cities or different countries. But since the basic problems are essentially the same and since most of the important index numbers published by the Government and private research organisations refer to data collected at different times, we shall consider in this chapter index numbers measuring change relative to time only. However, methods described can be applied to other areas also. USES OF INDEX NUMBERS Index numbers are indispensable tools of economic and business analysis. Their significance can be best appreciated by the following points: (9 They help in framing suitable policies. Many of the economic and business policies are guided by index numbers. For example, while deciding the increase in dearness allowance of the employees, the employers have to depend primarily upon the cost of living index. If wages and salaries are not adjusted in accordance with the cost of living, very often it leads to strikes and lock-outs which in turn cause considerable waste of resources. The index numbers provide some guideposts that one can use in making decisions.ee — of busine. evaluation SS eyesore fields also where videl yer ak of population s al je nl nay Spe ‘Though Mn a there 18 large ciologists ™ which are essentially and economic cé For example, ith that of an average Index numbers are useft Cire intelligene ‘isplay changes in the indices: earcholvet ring a person's Pa a h organisations have index numbers comp: thorities P! tiveness of school systems, cctivene: er age: health aut ducations nena Pe ricepital facilities Sein the effectiverumbers are most widely devised formulae to measure aH” «, since index Melt ne series SO formed (i They reveal trends and dena a period of time. ymenon under study. For used for measuring changes over 2 PEEP e phenomend” for the last 8-19 enable us to study the general re OF imports orward tendency, Le., they example, by examining index nlm tt rowing an UP) index numbers of years we can say that our imports 21% by examining the (ewhy caterers are rising year after year. Similarly, etc., for the last few Pat a n industrial production, business activity. d business activity. By ining conclude about the trend of production aft on can draw very important the trend of the phenomenon under study. We Tort ie to the effect of conclusions as to how much change is t etc. Thus index numbers are Seasonality, cyclical forces. irregular forces. °C. Oo highly useful in studying the general business conditions: ity index (ii) They are important in forecasting fu d present workings of our humbers are useful not only in studying the past and present workings 01 ov economy. but they are also important in forecasting future onomictequgy. Index numbers then are often used in time series analysis. the historical study of long-term trend, seasonal variations and business cycle development, so that business leader may keep pace with changing economic and business conditions and have better information available for decision-making purposes, (tv) Index numbers are very useful in deflating’. Index numbers are highly useful in deflating, ie., they are used to adjust the original data for price changes, or to adjust wages for cost of living changes and thus transform nominal wages into real wages. Moreover, nominal income can be transformed ilpibens come and nominal sales into real sales through appropriate index numbers. 518 al researc! Classification of Index Numbers Index numbers may be classified economics and business the classific: value; and (4) special purpose. in terms of wi ‘ations are : (1) hat they measure. In Price; (2) quantity; (3) of Simpson “ Statistical devices, “yi2°% numie the pulse of the of inflationary or INDEX PRO! Befor follov - cons meat and cons inde shou upp! wou Oth: inch the T bou avai Ea nur com bas peri poir Wi bas wal ist whi pro con to spiINDEX NUMBERS 51L PROBLEMS IN THE CONSTRUCTION OF INDEX NUMBERS Before constructing index numbers a careful v Bele See eful thought must be given to the 1, The Purpose of the Index At the very outset the purpose of constructing the index must be very clearly decided what the index is to measure and why? There is no all-purpose index. Every index is of limited and particular use. Thus, a price index that is intended to measure consumers’ prices. must not include wholesale prices. And if such an index is intended to measure the cost of living of poor families, great care should be taken not to include goods ordinarily used. by middle class and upper-income groups. Failure to decide clearly the purpose of the index would lead to confusion and wastage of time with no fruitful results. Other problems such as the base year, the number of commodities to be included, the prices of the commodities, etc., are decided in the light of the purpose for which the index is being constructed. The problem of the scope of the index, ie., the field covered by the index, is bound up with the purpose of the index and the data available. The data available, or rather the lack of them, may necessitate the modification of the purpose. election of a Base Period Whenever index numbers are constructed \Grreference is made to some™base period. The base period of an index number (also called the reference period) is_the period against which comparisons are made. It may be a year, a month or a day. The index for base period is always taken as 100. Though the selection of the base period would primarily depend upon the object of the index, the following points need careful consideration of base period: () The base period should be a normal one. The period that is selected as base should be normal, ie., it should be free from abnormalities like wars, earthquakes, famines, booms, depressions, etc. However, at times it is really difficult to select a year which is normal in all respects—a year which is normal in one respect may be abnormal in another, To, solve this problem an average of a number of years, 3 or 4 (preferably covering one complete cycle), may be taken as the base. The process of averaging will reduce the effect of extremes. Thus the average of the period from 2006 to 2008 may be considered normal: whereas no’ individual year in that span can be considered normal. , ' (i) The base period should not be too distant in the past. It is desirable to have an index based on a fairly recent period, since comparisons with a familiar set of circumstances are more helpful than comparisons with vaguely remembered conditions. For example, for deciding increase in dearness allowance at present there is no advantage in taking 1980 or 1990 as the base; the comparison should be with the preceding year or the year after which dearness allowance has not been revised. (ii) Fixed base or chain base. While selecting the base a decision has to be made as to whether the base shall remain fixed or not, ie., whether we have a fixed base or chain base index. In the fixed base method, theseer year or the period 0! Constant for all times. On the 01, Prices of a year are linked wit the fixed year. Naturally t ARbase the fixetcting the in than what is obtained by the depend upon the purpose of 0’ 3. Selection of Number of Items structing ‘The items ose for whi should be determined by the Pere construct Every item cannot be includ hence one has to select a sample. index it is impossible to include necessary to decide what commodi be selected in such a manner thi habits and customs of the people fo consumer price index for working class, For example, each an ties to inclu at they 1 whom nd. f the pr 0 se he chain par method. nd every de. are represent the index is items A deci: gtaTISTICAL METHODS ; ices are related is other Pr pase method the he cham year and not with ese ves a better picture givwever, much would \dex: included dex i in an index is constructed. >in ich the Mindex number and an u nile constructing a price » wh’ommodity. Hence it is \dities should .’ The commo sentative of the tastes, meant. Thus in a like scooters, motor cars sion must also be made ualities. Here we refrigerators, cosmetics, etc., find no place. . ige . ics, find no F s cluded, the more on the number of commodities to be inch should note that the larger the number o representative shall be the index but at thi be the cost and the time taken. deciding the number of commodities. Thus larger number of commodities shall have to b' specific purpose index as the index number of the pri industrial raw materials. f commodities in e same time. the greater shall ‘of the index shall help in The purpose h in a general price index a be included as compared to a »s of foodgrams or It is also necessary to decide the grade or quality of the items to be included in the index. Index numbers shall give wrong result if at one time one set of qualities is included and at another time another set. To avoid confusion about qualities it is desirable that as far as possible, only standardized or graded items are included so that they can be easily identified after a time lapse. 4. Price Quotations After the commodities havi f e been selected, problem is to obtain price quotations for these commodities. It ae a known fact that prices of many commodities y. fr even from shop to shop in the same market. tse ene ace, 2 Place-and price quotations from all the places where It is impracticable to obtain @ commodity is dealt in. A selection must be made of representativ e pl: a should be those which are well known for ment Persons. These places commodity. After the places fro obtained is decided, the next thin, m where thi is to a ig PPpoint some person or institution as and whi trading for that particular ~ Price quotations are to be ke pr Tt al tr u n fiTHODS ed is 1 the with, sture ‘ould dex sted. and rice tis vuld tes, aa urs, ade we ore, all in a a or od of n or e DEX NUMBERS, 521 kg. for one rupee. The re “hie aoung als ™ethod is free from confusion and is generally cision mus ane mane ci Be made as to whether the wholesale prices or retail prices Se Shuice Would depend upon the purpose of the index. 7 Bis See Pie index the wholesale price shall not be representative at all. is prices of certain commodities are controlled by the government then It ts these controlled prices that should be taken into account and not the black market prices which may be much higher, mean. median, mode, geometric mean or ni should be used for conatrueHng|thelindex (Median mode: ae EMea rer eee Traceeegea used in the construction of index numbers. Basically a chotce has tobe made between arithmetic mean and geometric mean. ‘Theoretically speaking, geometric mean is the best average in the construction of index numbers because of the following reasons : (9 in the construction of index numbers we are concerned with ratios of change; (i) geometric mean is less susceptible to major variations as a result of violent fluctuations in the values of the individual items ; and (ii index numbers calculated by using this average are reversible and therefore, base shifting is easily possible. The geometric mean index always satisfies the time reversal test. Despite theoretical justification for favouring geometric mean, arithmetic mean is more popularly used while constructing index numbers. This is for the reason that arithmetic mean is much more simple to compute than the geometric mean. However, wherever possible in the interest of greater accuracy, geometric mean should be preferred. It is gratifying to note that with the growing use of electronic computers geometric mean is becoming more popular in constructing index numbers. 6. Selection of Appropriate Weights" The problem of selecting suitable weights is quite important and at the same time quite difficult to decide. The term ‘weight’ refers to the relative importance of the different items in the construction of the index. All items are not of equal importance and hence it is necessary to devise some suitable method whereby the varying importance of the different items is taken into account. This is done by allocating weights. Thus, we have broadly two types of indices— unweighted indices and weighted indices. In the former case, no specific weights. are assigned whereas in the latter case specific weights are assigned to various items. It may be pointed out here that no index is unweighted in the strict sense of the term as weights implicitly enter in unweighted indices because we are giving equal importance to all the items and hence weights are unity. It is, therefore, necessary to adopt some suitable method of weighting so that arbitrary and haphazard Weights may not affect the results. There are two methods of assigning weights : (9 implicit, and (i explicit. In implicit weighing, a commodity or its variety is included in the index a Number of times. Thus, if wheat is to be given in an index twice as much Weight as rice, then two varieties of wheat against one of rice may be included ——= Weighting is the term used to describe conscious effort to assign to each commodity an influence that. in the final result, is proportionate to its relative ‘mportance.” Richardson.= ward 522 ing some oul weighting en. When the caseyofiexpllcltn is) VON Nora In the series. On the other hand. iM teins in ia ‘what do we weigh ? s p yf the va ed the i e use? conomic import roduction figures: evidence of importance 0! explicit weights are assign (i What type of weight do we (0 In order to bring ou involved the weight can be P ue weights. A quanity distribution figures. antity weights and valu mmodity produced, (i Weights are of two types : dUaNUS.C amount of comint "on the other weight, symbolised by q. means Oo iod. A value WED consumed’. distributed, or consumed in some time Period“ Gictriputed’ OF CONSE « hand, combines price with quantity ‘produces. 7 where p st Value is in terms of rupees and is symbolised PY aie price and q for the quantity. me or value weights. If Now the question is whether to choose qua) sting index, then quantities a the aggregative method is used while constructié 1" ive the same units, used as weights because price times quantity Wi ng price relatives quantity namely, rupees. On the other hand, in averaging PriCe Te rcentages figures cannot be used. It is for the reason that if We MUTE AT nits: for by quantities expressed in different units, we get res! 'd percentage times kg. example, percentage times tonnes will give tonnes and perceninsy (re ee will give Ikg. Such figures cannot be used in computation. ECT Bete et are multiplied by value figures, which are always expressed in TPT TS At answer in rupees only. Hence the statistician will use q as a weg Ne method of aggregating actual prices and must use px q as a weig) method of averaging price relatives.* 4 ab Another problem in connection with weights is that of deciding whether the weight shall be fixed or fluctuating. Since the relative importance of the different items does not remain the same for all times it is logical to vary the weight from time to time. Such an index would give better results. However, when fluctuating weights are used one must be very careful in interpreting the index because not only changes in prices but also changes in weights are affecting the index. et . questions a of the commodities nce sumption figures or consul y weights 7. Selection of an Appropriate Formula have been devised for constructing the index. The problem very often i: that of selecting the most appropriate formula. The choice of the formuln would depend not only on the purpose of the index but also on thee mae available. Prof. Irving Fisher has suggested that an appropriate (qty that which satisfies time reversal test and factor reversal ra ae cally, Fisher's method is considered as “ideal” However, from a practical point of view there ave eeu’ index m this index which which shall be discussed later “an set oneal Bare ule AMO ARenNBe eared asthe best under ai circumatenaen e is Imowedge of the characteristics of differecenStanCes- a diseriminating investigator Will choose technical. meio. OTM data and ap] to his purposes. adapted to his is simple to solve in ise between th A large number of formulae Practice and the fj inal in eoretical Standards and nes Magnitudes may have to these weigh Ls. of index numbers,iNDEX NUMBERS, METHODS OF CONSTRUCTING INDEX NUMBERS ‘A large number of furmulae have been devised f structing index numbers. Broadly speaking, they can be grouped uncer two heads: (a) Unweighted indices; and {b) Weighted indices, in the unweighted indices weights a : re not expressly assigned whereas in tne weighted indices weights are assigne ees a ee types may be further divided under tw heasee rious tems: Fa (o Simple Aggregative, and (i Simple Average of Relatives. ‘The following chart illustrates the various methods : Index Numbers Unweighted Weighted i Simple Simple Average Weighted Weighted Aggregative _ of Relatives Aggregative Average of Relatives Unweighted Index Numbers 1. Simple Aggregative Method This is the simplest method of con- structing index numbers. When this method is used to construct a price index, the total of current year prices for the various commodities in question is divided by the total of base year prices and the quotient is multiplied by 100. Symbolically : Po. = oe x 100 ¥ P; = total of current year prices for various commodities. ¥ Pp = total of base year prices for various commodities. This method of constructing the index is the simplest of all the methods. The steps required in computation are: () Add the current year prices for various commodities, ie., obtain ¥ Py. (i) Add the base year prices for the same commodities, ie., obtain ¥ Py. (it) Divide S P; by E Po and multiply the quotient by 100. lllustration 1. From the following data construct an index for 2008 taking 2007 as base : Commodity Price in 2007 (Rs.) Price in 2008 (Rs.) A 50 70 B 40 60 c 80 90 D 110 120 E 20 20graTISTICAL METHODS 524 Solution. ae ‘Commodity moons 360, 499 = 120 300 re ices of me am ase in the prices J 2007, in 2008 there is a net Ino This means that as compare , ‘commodities included in the index to the extent of 20%. simple aggregative Limitations. There are two main limitations of the simP eae ‘ rt a big can. exe + The units used in the price or quantity quotations influence on the value of the index. ve nce of the + No consideration is given to the relative importan commodities. 4 a tee Zhis index is based on the assumption that the various terms and their Bie re aioted| {nthe seme unt. the unit of measurement is diReheagiy different items, the index shall produce vastly divergent results. To i ae et us consider the problem of calculating an index comparing the 2008 cost of construction with that of 2007, and let us include only two items, the cost of labour and the price of cement. 2007 2008 Average hourly wage to be paid to construction workers Rs. 2.50 Rs. 2:80 Price of cement per bag Rs. 60 Rs. 65, =P, = 67.8, DPo= 62.5 8 Po, = G95 * 100 = 108.48 per cent, Now suppose instead of hourly wage we week of 50 hours we can rewrite the figures a: Average weekly wage take the weekly wage. Taking a ~ Rs, 125 Price of cement per bag Rs. 60 Rs. 140 Now XP, = 205, ¥Po = 185 Recs 205 Po, = Tg5 * 100 = 110.81 per cent, One can see at once that when we used wee to guntiek increased from 108.48 to 110.81. Matinet ‘o suit one's requirements by quoting per kg. rather thas 2°¢ thus possibie maund rather than per quintal, and so on ara” cl Maubulated in this manner can hardly be qc¥qdex Wi Moreover, equal importance is sé value edie ance an be given to alll the it, objective mea, these reasons that the simpl items which j, SS degree of acceptance, © Sezregative index hy me aa my never gaineq any great pt lusBai — + if a 10 5 9 : 4x 100-60 "200d 5 2 100 180 2000 5 $x 100= 100 2005 s 42 x 100=200 ao S ern ag “ 14 x 100 = 220 2002 7 2x 100= 140 2007 2003 8 ex 100 = 160 ‘ ‘ee ee this method is 2. Simple Average of Price Relatives ee Hon relatives are obtained ‘i d then avtral value, be, x ant n res of cent i harmonic mean. When formula for com- used to construct a price index, first for the various items included in the Dee relatives is obtained using any one of the metry arithmetic mean, median, mode, geometric MeN TT the arithmetic mean is used for averaging the relatives puting the index is : \ Pi rw 100] z Pui ) Poe Ne where N refers to the number of items (commodities) whose price relatives are thus averaged. a Although any measure of central value can be used to obtain the overall index, price relatives are generally averaged either by the arithmetic or the geometric mean. When geometric mean is used for averaging the price rela- tives the formula for obtaining the index becomes tr, > log} — x 100 *1Po | log Pox = or EP where p=t 100 0 OR iB log a x 1) Po) = antilog eee vig a= antilog Ss tes P Other measures of central value mm relatives. SeaPeeea common use for averaging * A price relative is the ratio of the pri Period to its price in another period called ki ‘eee commodit pe If po and pn denote the commodi , tiod or period respectively, then by definition: Pi’ GUFINE the base: pekTenCe period. ty in a given ase Period and givenis INDEX NUMBERS mustration 3. From the following data construct an index for 2007 taking 2006 as base by the average of relatives method using (a) arithmetic mean, and (b) geometric mean for averaged relatives ‘Commodity E Price in 2006 (Rs.) 50 40 80 110 20 Price in 2007 (As.) 70 60 90 120 20 Solution. (a) INDEX NUMBERS USING ARITHMETIC MEAN OF PRICE RELATIVES _ ‘Commodity Price in 2006 (Rs.) Price in 2007 (Rs.) Price relatives fo a Ps 100 pins = pore A 50 70 140.0 B 40 60 150.0 c 80 90 1125 D 110 120 109.1 | ae 20 a coe 100.0 me ote 2 Epp X 100=611.6 2x 100 Por = "5 — = 122.82 ___ (INDEX NUMBER USING GEOMETRIC MEAN OF PRICE RELATIVES ‘Commodity Price in Price in Price Log P 2006 2007 Relatives pee iia aioe Li) SV hella - A 50 70 140.0 2.1461 B 40 60 150.0 2.1761 c 80 90 112.5 2.0512 D 110 420 109.4 2.0378 aE 20 20 100.0 2.0000 Blog P= 10.4112 Por = Antilog lara = Antilog [SS | Antilog 2.0822 = 120.9 Although arithmetic mean and geometric mean have both been used, the arithmetic mean is often preferred because it is easier to compute and much better known. Some economists, notably F.Y. Edgeworth, have preferred to use the median which is not affected by extreme values. Since the argument is important only when an index is based on a very small number of commodities, it generally does not carry much weight and the median is seldom used in actual from the data given below : RATE PER RUPEE Commodities 1 35 kg. 3 kg. | Practice. Illustration 4. Prepare Index Numbers of price for three years with average price as base A B G ist_year 10 kg. 4 kg. 3 kg. 2nd year 9 kg. 3rd year 9 kg. 3 kg. 2ikgerage price and jerermine 2 (in rupees) =100) Year 528 into Solution. Convert first the prices I ray E with this average price as base comput mes “| Unit | Average | a ” 115 110 330 110 Note. P indicates Price Relatives ai For 1st year 10 kg. of commodity A costs Re. 1 = 4 Hence 40 kg. of commodity A will cost 7p x 40= RS: ‘Similarly other prices are obtained. 43's the average of 4, 44 and 4.4, i. ars the respective prices of the three yo i ages over the previous Merits This method has the following two advantage method : 7 i ce is given + Extreme items do not influence the index. Equal importan\ a to all the items. : ie + The index is not influenced by the units in which prices are quote or by the absolute level of individual prices. Relatives are pure numbers and are, therefore, divorced from the original units. Conse- quently, index number computed by the relatives method would be the same regardless of the way in which prices are quoted. This simple average of price relatives is said to meet what is called the units test.* Limitations Despite these merits this method is not very satisfactory be- cause of two reasons: selection of an appropri average. The use of the arithmetic mean is considered ac quatre Sometimes because it has an upward bias. ‘Th i mean involves difficulties of computation, Other aveneesy ‘a never used while constructing index numbers Ss, are iauuoay + The relatives are assumed to have equal i import: is’ 18 pal kind of concealed weighting system that te ane ug eau economically some relatives are more important’ than a me Te ers. Weighted Index Numbers The ‘so-called unweighted index numbe; in the true sense of the term. They ae ccus items included in the index and as such the ts implicit rat than ality weighted, ul ‘arlier in case of aby, changing the Tealistic in Most of Number Formutae a t itime [NDEX NUMBERS —_ ina tne cases. Construction of useful index numbers requires a conscious effort a fo assign to each commodity a weight in accordance with its importance = in the total phenomenon that the index is supposed to describe. Weighted index numbers are of two types : « Weighted Aggregative Indices, and « Weighted Average of Relatives weighted Aggregative Indices These indices are of the simple aggrega- tive type with the fundamental difference that weights are assigned to the bi yarious items included in various methods of assigning weights and con- sequently a large number of formulae for constructing index numbers have been devised of which some of the more important ones are: + Laspeyres method + Paasche method + Dorbish and Bowley’s method s + Fisher's ideal method + Marshall-Edgeworth method, and a + Kelly's method All these methods are named after the persons who have suggested them. a e () Laspeyres Method. The Laspeyers Price Index is a weighted aggregate price index, where the weights are determind by quantities in the base period. The formula for constructing the index is : =P1d0 Pogo Po.= x 100 Steps : + Multiply the current year prices of various commodities with base year weights and obtain Dpiqo- + Multiply the base year prices of various commodities with base year weights and obtain Dpodo- + Divide Spiqo by Zpodo and multiply the quotient by 100. This gives t | us the price index. i Laspeyres Index attempts to answer the question : “What is the change in | aggregate value of the base period list of goods when valued at given period ] Prices 2" This index is very widely used in practical work. The primary disadvantage of the Laspeyres method is that it does not take | into consideration the consumption pattern. The Laspeyres index has an upward bias. When prices inerease, there is a tendency to reduce the consumption of higher priced items. Hence, by using base year weights, too much weight will be given to those items which have increased in price the most. Similarly, weg prices decline, consumers shift their purchases to those items which : cot the most. By using base period weights, too little weight is given to ‘hose items which decrease most in price, again overstating the index. This method was devised by Laspeyres in 1871 and that is why it is called Laspeyres Method.oy ‘I eighted ager... x is a Ww - Hay ice inden ve by quantities in the Steps. rice: + Multiply current year P! eunie and obtain Spi 41- ; price: + Multiply the base year prices 4 Weights and obtain a y the quotient by 100. ; iply Racecep teen Dy 20 a a pestion : “What would be the vatye Jn general this formula answers the question in ted prices 7 of the given period list of goods when he Index in practice 1s that re The difficulty in computing the ce ach year of each period. adding the'data colecton expense in the nreparation of the index, For thls reaspr the data collection expense in the prep: iy in practice where the number ot the Paasche index is not used frequently in pra SE. ik res index measures Comparison of Laspeyres and Paasche methods. Laspey’ r change in a “ved market basket” of goods and services. The same quantities are used in each period. The Paasche index continually updates the quantities to the levels of current consumption. These two approaches tend to Produce poposite extremes in index values computed from the same data. From a ractical point of view, Laspeyres index is often preferred to Paasche’s for the P Pe Pt Simple reason that in Laspeyres index weights (qo) are the base year quantities and do not change from one period to the next. On the other hand, the use of Paasche index requires the continuous ity weights for each Period considered and in most cases these weights are difficult and expensive peetiaia, i spice countries index numbers’ are constructed by using An interesti is generall ie rent ities with cu Year s commo s of variou odities with curren, ious comm 5 of vario year quantities we wil] be giving to increased the most and the rita a nae When the prices 80 down, consu; items which have declined in the numerator of weight te th again be too large. their numerator will SS on 800ds when Ng produces an a downward €s index must Ie WI ‘ho first Used it in 1874. INEEMHong, Tegate n the Year Trent value ‘ised 1g to son, T of ares ities ties uce na the ties 2 of ach sive ing rer he res, for ke ts nt al in cd st INDEX NUMBERS 531 unless drastic changes have taken place between the base year and the given year. the difference between Laspeyres and Paasche's will generally be Small and either could serve as a satisfactory measure. In practice, however, the base year weighted Laspeyres type index remains the most popular for reasons of its practicability. The Paasche type index can only be constructed when up-to-date data for the weights are available. Furthermore, the price index of @ given year can be compared only with the base year. For example. Jet Pog = 100, Pog = 130 and P95 = 145. Then Pog and Pog are using different weights and cannot be compared with each other. If these indices had been obtained by the Laspeyres formula they could be compared because in that case the weights are the same base year weights (qq). For these reasons, in practice the Paasche formula is usually not used and the Laspeyres type index remains most popular for reasons of its practicability. (ii) Dorbish and Bowley's Method. Dorbish and Bowley have suggested simple arithmetic mean of the two indices (Laspeyres and Paasche) mentioned above so as to take into account the influence of both the periods, Le., current as well as base periods. The formula for constructing the index is L+P Po = “+ where L=Laspeyres Index, P= Paasche Index pido Fe rpg or Po, = 2Poto_Z oa > EPoM 109 (iv) Fisher's ‘Ideal’ index. Prof. Irving Fisher has given a number of formulae for constructing index number and of these he calls one as the ideal’ index. The Fisher's Ideal Index is given by the formula : Zpido , Spid eet Pai = 100 or Po, = \LxP Fad ea iilin utinnertee an It shall be clear from the above formula that Fisher's Ideal Index is the geometric mean of the Laspeyres and Paasche indices, Thus, in the Fisher's method we average geometrical formulae that err in opposite directions. The above formula is known as ‘Ideal’ because of the following reasons : () It is based on the geometric mean which is theoretically considered to be the best average for constructing index numbers. (i It takes into account both current year as well as base year prices and quantities. (i) It satisfies both the time reversal test as well as the factor reversal test as suggested by Fisher.* (iw It is free from bias. The two formulae (Laspeyres and Paasche’s) that embody the opposing type and weight biases are, in the ideal formula, crossed geometrically, ie., by an averaging process that of itself has no bias. The result is the complete cancellation of biases of the kinds revealed by time reversal and factor reversal tests. ——-s For proof please see under “Tests of Adequacy of Index Numbers”.STATISTICAL METHODS use it is excessively a2 yute becal It is not, however. @ practical Jone seagene segment of oe nee arena As ty The data, particularly will continue to laborious. The Om\in practice, statisticians 7) readily availabl Although perhaps less exact, <<: or d also both ») Marshall-Edgeworth Method. In en ee | anaidered. The formula for ©) Nell-as base year prices and quanti c ting the Index is : constructing eerie Pou = (qo +41) Po 1 opening the brackets ange z _ Dprgo+® Prd, 199 01" E podo + © Pod It is a simple, readily constructed measure, giving a Ve! tion to the results obtained by the ideal formula. hod, ‘Truman L. Kelly has suggested the following formula ry close approxima- (vd Kelly's Met for constructing index numbers : Po, = P44 x 100 > pod Here weights are the quantities which may refer to some period, not necessarily the base year or current year. Thus the average quantity of two or more years may be used as weights. if in the Kelly's formula the average of the quantities of two years is used as weights, the formula becomes : Poi = cae x 100; where q= Boo Similarly, the average of the quantities of three or more years can be used as weights. This method is known as fixed weight Mcaregutie index and is currently in great favour in the construction of index number series. An important advantage of this formula is that like Laspeyres index it does not demand yearly changes in the weights. Moreover, the base period can be ch ee pecesattating corresponding change in the weights. This is ecauuse the construction of appropriate quanti 7 general purpose index usually requires a considerible. amie G Rae eights can thus be kept constant until new census (or other eure oe become available to revise the index. other survey) data IMlustration 5. Construct index numbers of pri 1. Laspeyres method, Of Price from the following data by applying ; 2. Paasche method, 3. Bowley’s method, 4. Fisher's Ideal method, and 5. Marshall-Edgeworth method. ‘Commodity (a alert tone e006 = ee, ae Quantity 6 6 5 5 BI 10 13 INDE? saa ~~V1-45574 x 100 = 1.2065 x 100 = 120.65 yALUE INDEX NUMBERS fa si , The value of a single commodity is the product of its price and quantity. jus a value index V is the s ae sum of the values of Sat Bees Foren erereuirer aa ke = 2Pid1 vs pogo * 100: V = Value index where zpid = Total value of all commodities in the given period, and z 2Podo = Total value of all commodities in the base period. peetac aa cases the value figures are given the formula can be stated ys | Pi uy | in which V stands for value. : In this type of index both price and quantity are variable in the numerator. Weights do not have to be applied, since they are inherent in the value figures. A value index therefore is an aggregate of values. It measures the change in actual values between the base and the given period. The vaule index is not in wide use, although because of the unsatisfactory | nature of price and quantity indices, it has been occasionally suggested that they be replaced by the value index. The temptation, however, must be resisted, since the concepts of price level and quantity level answer questions that cannot be answered by the value level. Furthermore, an aggregate of values may be viewed as the product of a price level and a quantity level. The division of an aggregate of value into its price and quantity factors may be arbitrary. but this arbitrainess need not create any confusion of thought as long as our concepts of the two factors are consistent. The test of consistency is that the product of the price and quantity indices must produce the value index. TESTS OF ADEQUACY OF INDEX NUMBER FORMULAE Several formulae have been suggested for constructing index numbers and the problem is that of selecting the most appropriate one in a given situation. The following tests are suggested for choosing an appropriate index : + Unit Test.* Time Reversal Test + Factor Reversal Test. * Circular Test. * Freund and Williams : Modern Business Statistics.bi oad a for constructing ap, 540 formula wires that bs which, ee 4. uate rose The unit test FEMS units ST nye (unwelgne tae eve index should be independ’ Cont for UO" chapter satisty and quantities are quoted. ET in tl cciveeen f formulae disc e a careful study of index all other fo} mad and has suggesteq as g Fisher Eber 2. Time Reversal Test Prof. ii x 0 Fea arererapessié for compuesto indica various tests to be applied to aY Ti" or these he satisfactory. The two most impor yen method will work saesraetoe actor reversal teSt~ ne whether a SVEN tthe testis Bae ie factor ror teat to determine WV yards Ot at sate soi avai be a bgt ga in H ; See ee one point of comparison ne data for any two years eee ee pen as bases in other words, WHC °° versed, the two index See ate pw the same method, but with the Pav” TO at their product is numbers secured should be reciprocals of €aCh Gone sed : tunity. Symbolically, the following relation should O° Pot x Pio = oe ee an Be Ge index for time “1” on time “0! a base ana Bt ie index for'time “0” on time “I” as base. If the product is not MaDe eter is said to be a time bias in the method. Thus if from 200: Hl a oS) he BHeetoe wheat increased from, Re. 460 to Rs. 560,perjquintal tHe geass age 02 and the price in 2002 whether or not it is the time reversa] 2000 should be 1335 per cent of the price in 20 should be 75 per cent of the price in 2003. One figure is the reciprocal of the other; their product (133. x -75) is unity. This is obviously true for each individual price relative and, according to the time reversal test, it should be true for the index number. The test is not satisfied by Laspeyres method and the Paasche method as can be seen below : When Laspeyres method is used aie =Pido XPoqo _ 2poa. _ 2pigo . Lpom Pio = Spar’ Po. x Pig = Src x Shiai #1 and the test is not satisfied. When Paasche method i used Pid Pogo + Pyio==—— Ha ope a pigo 1 ,, 2PoA0 Po. X Pig = Soe x Poly P10= Soa * Spido* and the test is not satisfied. ‘There are five methods which do satisfy the test ; r Ideal formula. of price relatives, an of m OF Price relatives if we use fixed sNDEX? Let Pro chi si 3. F fact qua wor two inte ie. oth in of the ye th be ve In ir iyypeX NUMBERS Zpod | Lpodo _ Zein “ Epiqo— since Poi * Pio = 1, the Fisher's Ideal Index satisfies the test. v=1 3. Factor Reversal Test Another test su factor reversal test. It hoids that the product of 2 price index and’ the quantity index should be usual to the corresponding value index. In the words of Fisher, “Just as each formula should permit the interchange of the wo times without giving inconsistent results, so it ought to permit interchanging the prices and quantities without giving inconsistent result, ie. the two results multiplied together should give the true value ratio.” In other words, the test is that the change in price multiplied by the change in quantity should be equal to the total change in value. The total value of a given commodity in a given year and the product of the quantity and the price per unit (total value =p xq). The ratio of the total value in one Prd Pod) year to the total value in the preceding year is If from one year to the next, both price and quantity could double, the price relative would be 200, the quantity relative 200, and the value relative 400. The total value in the second year would be four times the value in the first year. In other words, if p; and po represent prices and q; and qo the quantities in the current year and the base year, respectively, and if Po, represents the change in price in the current year and Qo; the change in quantity in the current year, then Pee Ont =p OLS $01. Stp0g0 If the product is not equal to the value ratio, there is, with reference to this lest. an error in one or both of the index numbers. } The factor reversal test is satisfied only by the Fisher's Ideal Index. vo Ypido , EPin o1= V'Spoqo =Podi Changing p to q and q to p _ {Sapo , Ea.P1 901 = N'Saop0 * E40P Spigo Spin , Sapo , LaiPr Por x 91 = VS pogo Spo E4oP0 EdoP1 _4[Gpia? _ P01 = Vispoa)? — 2P0% Proof :tisfied by the Fisher's sal est 18 eversal te 5 equally well for EpIM the factor Teve ta serves ee == mul .s of pl » the Sine’ Por 901° Spode) urge, that the Orting indices OY formula, Ideal index. This means. of OFS, for CONS ind q iM UM metic mean, constructing indices Owed by interchanBine dry indices factor reversal quantity index being derived by Inter Pchementary Mey 0 ae None of the simple or weighted f0r™8 TF" tne requiteM" posed by the f harmonic mean, geometric DO ng reatl ong ction ‘of many highly test. It is thus obvious that in the col reversal test compel its being ignored !" ge are no good logical reputable index numbers: ect argue that the re these tests. For Some authorities on the s ae pape reasons for claiming that Ca ears time reversal #24 in OR ce oul s a collection of goods incded in Fb, is diflerent from Neonsistent results. against q)) and, therefore, one could hardly hop’ of the index number adequacy Bee 4. Circular Test Another test of the ade ¢ of index numbers 5 what ts no Ir in to years, Dut to the formula is what is known as ‘circular test’ interest attaches not merely to a comparison ie eet it is frequently measurement of price changes over a perio d to meet this test if, for desirable to shift the base. A formula is said to mece Ca the 99g example. the 2003 index with 1998 as the base is A00. aul oe oo index with 1993 as the base is again 200, then the meee as the base must be 400. Clearly, the desirability of this property is that it enables us to adjust the index values from period to period withou referring each time to the original base. A test of this shiftability of base is called the circular test. 4 This test is just an extension of the time reversal test. The test requires that if an index is constructed for the year a on base year b, and for the year b on base year c, we ought to get the same result as if we calculated direct an index for a on base year ¢ without going through b as an interthediary. Symbolically, if there are three indices po,, pz and pyo the circular test will be satisfied if - Pol P12 X Pag ='1 The Laspeyres index does not satisfy the test a: aa is can be seen from the If the three years are 0, 1, 2 the index by Las 2pigo , Eo x SPO peyres method will be e fixed weight ing : e will get : it a iti prog serega Ve method, we will get :ETHODS isher's ell for 8, the mula, nean, versal factor lighly rgical For med Toas aber bers the ntly for 993 993 hat out ase hat on. dex vill he d. at INDEX NUMBERS s. An index which satisfies this test has the advantage of ing the computations every time a change in the base year ae peiaae ‘Such index numbers can be adjusted from year to year without referring each time tothe original base. The circular test is not met by the ideal index or by any of the weighted aggregative with changing weignts. The test is met by athe ect Fond of rice relatives and the weighted aggregative fixed weights. The reason why the Laspeyres and Paasche index numbers and their derivatives, the Marshall-Edgeworth and the Ideal indices do not meet the circular test is that the weights in these index numbers depend on the periods between which comparisons are being made. If these periods change, the weights change. For example, if the base period is taken as period 2 rather than period 0, the weights of Laspeyres index are no longer qo but qo. Karmel has pointed out that although it may seem reasonable to argue that if a price index between periods 0 and 1 has risen to M and between periods 1 and 2 to N, then between periods 0 and 2 it should have risen to MN, a moment's reflection will show that this requirement is not reasonable. An index number has meaning only in terms of the system of weighting adopted, and one may produce many numerically different but quite valid indices for comparing two periods. The weighting system used in Pio (Laspeyres) is the same as that in Po; (Laspeyres), but different from that in Py2 (Laspeyres). Consequently, the increase M is an increase in something different from that in which N is the increase. The product MN is, therefore, a mixture, the exact meaning of which is not clear and which could not be expected to equal a direct comparison between periods 0 and 2. Illustration 12. Calculate Fisher's Ideal Index from the following data and prove that it satisfies both the time reversal and factor reversal tests: 2006 2007 Commodity Price Expenditure Price Expenditure A 8 80 10 120 B 10 120 12 96 G 5 40 5 50 D 4 56 3 60 = 20 100 25 150 = (B.Com., Madras Univ., 2005) Solution. CALCULATION OF FISHER'S IDEAL INDEX ‘Commodity 2006 2007 Po % pr pido Pogo pigs pom a 8 10 OMe Te 700 80 120 ~«96 8 10 12 12 8 144120 96 80 c 5 8 5 10 40 40 50 50 D 4 14 3! 820) 42 56 60 80 E 20 5 25 6 125 100150120 “pig Deg = Leg = pon =451_=396_—=476_—=426 a 20190, Zea = \V45t 76 Por="VSpoge “Sma * 100 Wage 426 * °° 72726 x 100 = 1128 x 100-1128. Time Reversal Test : Time reversal test is satisfied when Poi x Pio= 1 Zpog | Dpom _ 1/426 396 = Wspiagi “ Spi 476 * 451 Px