ISSN (Print) - 0012-9976 | ISSN (Online) - 2349-8846

A+| A| A-

Domestic Terms of Trade in a Three-Sector Framework

The domestic terms of trade issue in India has mostly been addressed by analysing the terms of trade between an "agriculture" and "industry" sector. An increasing share of the Indian GDP today however originates from a "third" service sector. The presence of a large service segment necessitates a switch from the two-sector to at least three-sector framework of terms of trade analysis. This paper estimates the terms of trade effect for these three domestic sectors across India and in 15 major state economies. The results confirm the crucial role played by services in determining the terms of inter-sectoral trade.

Domestic Terms of Trade in a Three-Sector Framework Analysis for All-India and States

The domestic terms of trade issue in India has mostly been addressed by analysing the terms of trade between an “agriculture” and “industry” sector. An increasing share of the Indian GDP today however originates from a “third” service sector. The presence of a large service segment necessitates a switch from the two-sector to at least three-sector framework of terms of trade analysis. This paper estimates the terms of trade effect for these three domestic sectors across India and in 15 major state economies. The results confirm the crucial role played by services in determining the terms of inter-sectoral trade.

SURAJIT DEB

I Introduction

T
he pattern of movements in domestic (sectoral) terms of trade (hereafter TOT) has generated substantial research interest in both academic and policy-making circles in India. The studies attempting to empirically estimate sectoral TOT movements have mostly been undertaken on the basis of examining the TOT between agricultural and industry (or nonagriculture) sectors in a two-sector framework. This is true of Thamarajakshi (1969, 1994), Kahlon and Tyagi (1980), Tyagi (1987), Mungekar (1992), Palanivel (1999) and also of two government agencies, viz, the Commission for Agricultural Costs and Prices (CACP) and Directorate of Economics and Statistics (DES), both in the ministry of agriculture. However, the Indian growth experience over the last three decades or so has led to a structural transformation, where an increasing percentage share of the domestic product seems to originate from service activities.1 Since the quantitative importance of service activities has grown in the all-India as well as in various state economies, it is desirable to acknowledge the role of services in domestic trade terms. To be specific, the structural tool of TOT analysis must be extended from the two-sector to a three-sector framework.

In this context, the main objective of the present paper is to examine the implications of formulating a three-sector analysis of domestic TOT through a case study of India. It may be crucially noted that the proposed alternative methodology uses an accounting framework, which takes into account the entire economy and not just two notional sectors. In addition to this, there are other advantages in using the alternative approach in TOT measurement. For one, it provides TOT estimates for all the major sectors of the economy by avoiding some of the sticky estimation problems associated with aggregate NBTOT calculations, such as the aggregation error and aggregation problems.2 Two, the same methodology can be used to generate consistent and comparable set of TOT estimates across different states in India on which there exists very little work today.

The article is structured in the following manner: Section II provides an outline of different concepts of TOT and resource flow measure. This is followed by a comparative analysis of the present TOT measure with the standard net barter TOT measure in Section III. The details on the sectoral definition and required database are described in Section IV. The pattern of TOT movements in the three-sector classification of the Indian economy, as well as in 15 major state economies are discussed in Sections V and VI, respectively. The time period of analysis is 1951-97 at the all-India level and 1971-96 at the level of individual states. Section VII concludes.

II TOT and Resource Flow Measures

Measures Based on NBTOT

The barter TOT and resource flow measures have remained important in examining the changes in relative price or income flow between two competing sectors. Various indices of TOT can be found in the work of Kindleberger (1956). These indices were originally discussed in the context of foreign trade between the developed and less-developed regions. In its simplest form, a TOT index generally referring to NBTOT is defined by:

PX

NBTOT = (1)PM

where PX and PM refer to composite price indices of exports and imports, respectively.

The two composite price indices are defined in such a way as to represent the aggregate price movements of exportable and importable commodities, which requires the respective commodity weights to be assigned on the basis of items actually traded by a given nation. The other index that evaluates the quantum of tradeables referred as the gross barter TOT (GBTOT) is defined by:

QX

GBTOT = (2)QM

where QX and QM stand for quantities (or values) of exports and imports, respectively.

Some other indices of TOT were also developed, e g, income TOT (ITOT), single factorial TOT (SFTOT) and double factorial TOT (DFTOT). The ITOT takes into account the effects of changes in prices of traded goods and the value of exports (or imports) as well. That is, while Dorrance (1950) defined ITOT as the value index of exports divided by the price index of imports,

QX PX

ITOT = (3.1)PM

Staehle (1951) specified ITOT as the value index of imports divided by the price index of exports, viz,

QP

MM (3.2)

ITOT = PX

While the former determines the capacity to import (or purchasing power of exports), the latter indicates the required import bill (or value of required exports) in a nation’s foreign trade.

On the other hand, the SFTOT and DFTOT take into consideration relative change in the productivity levels of export and import originating foreign sectors. These indices were developed to reflect improvements in commodity TOT that depend on productivity growth. The SFTOT is defined as the commodity TOT multiplied by an index of domestic productivity level. The DFTOT also takes into account the foreign productivity level and is defined by the commodity TOT multiplied by the ratio between domestic and foreign productivity levels. For empirical estimation, DFTOT is thus worked out as the ratio of unit value index of exports to that of imports after adjusting for the changes in their productivity levels. That is,

PX TX

DFTOT = (4)PMTM

where TX and TM denote productivity indices in the export and import sectors, respectively.

The TOT worked out for Indian agriculture mainly refer to the indices of NBTOT and ITOT. The basic methodology, pioneered by Thamarajakshi (1969), has involved the construction of composite indices of prices received and prices paid by agriculture for its traded goods. Subsequently, Kahlon and Tyagi (1980, 1983), Tyagi (1987), Mungekar (1992) and Palanivel (1992, 1999) have attempted empirical estimation of NBTOT and ITOT. The indices of ITOT were developed on the basis of Dorrance’s measure above. Palanivel (1992) also provides estimates of DFTOT by utilising total factor productivity indices in agriculture and manufacturing. Apart from these, the two government agencies, viz, CACP and DES in the ministry of agriculture also estimates the NBTOT between agriculture and non-agriculture, which are used for the purpose of devising the agricultural procurement (or support) prices.

Inter-Sectoral Resource Flow Measure

There is also a measure, which is related to NBTOT that estimates the flow of resources between the agriculture and nonagricultural sectors in an economy. Ishikawa (1967a, 1967b) used this approach to explore the direction and magnitude of net resource flow out of agriculture during the initial phases of development in Asian economies such as Japan, Taiwan, India and mainland China. Following this, several country specific attempts were made by Lee (1971), Mundle and Ohkawa (1979), Mody, Mundle and Raj (1985), Sheng (1993) and Karshenas (1995) and in the Indian context by Mundle (1977a, b, 1981), Mody (1979, 1981) and Palanivel (1992). In this approach, the net inter-sectoral resource flow (NIRF) out of agriculture is derived by estimating the balance of inter-sectoral commodity trade of the farm sector, i e, NIRF = R = M – E (5) where M, E and R stand for the current values of import, export and import (or export) excess, respectively.

The real NIRF is obtained in terms of real values of imports and exports. Ishikawa (1967a) defined this as:

ME Real -NIRF = —– –—– (6)PM PE

where PM and PE are the composite price indices of the import and export commodities, respectively. In Ishikawa (1967a), the real NIRF has been decomposed as:3

ME R 1 ⎛ PE ⎞ (6.1)

Real NIRF = −= + E ⎜⎜ −1⎟⎟ MEME ⎝ M ⎠

PPPP P

where (1/PM)R represents the NIRF in a visible form. The other part, (1/PE)E(PE /PM –1) has been considered to represent the invisible NIRF caused by changes in inter-sectoral TOT.

TOT Effect in NIA Framework

Another stream of systematic research has attempted to measure the “sectoral TOT effect” by using the national income accounting (NIA) principles. This line of approach, which began with the works by Stuvel (1956) and Rasmussen (1957) has attempted to adjust the sectoral income for changes in inter-sectoral TOT. The “TOT effect” accruing to a sector is defined by the income gain (or loss) due to changes in the prices of exports relative to imports in the inter-sectoral trade. Subsequently, Bjerke (1968, 1972), Olgaard (1966, 1981) and Derksen (1980) have used this framework in the context of examining TOT for the Danish and Dutch economies. Two different concepts of TOT effect can be distinguished from their exposition, viz, one that defines TOT gains from a production point of view and the other differentiates TOT gain from an income point of view. While the production gain evaluates a sector’s purchasing power over the GDP basket, the income gain is drawn from the purchasing power of final consumption goods. The TOT effects accruing to the j-th sector is defined as:

Gain (Production)j = 1 [X (P − P )] (7.1)

va. j va. j vaPva

Gain (Income)j = 1 [X ′(P − P )] (7.2)

va. j va. jDPD

where: X´va.j = sectoral gdp of the j-th sector at constant prices Pva.j = implicit price deflator for the j-th sector, i e,

Pva.j = Xva.j/X´va.jPva = implicit price deflator for the economy, i e, Pva = ΣjXva.j/ΣjX´va.jPD = price index of final demand (consumption plus investment) PDj = sector specific price index of final demand for the j-th sector.

Economic and Political Weekly April 29, 2006

This paper proposes, perhaps for the first time, to use the NIA framework so as to estimate the “TOT effects” accruing to agriculture, industry and services sectors in the all-India and in 15 major state economies. Some conceptual differences of the measure based on NIA framework from the commonly used barter TOT measure are discussed below.

III Comparison of Alternate TOT Measures

The barter TOT have been widely used in the analysis of variations in sectoral prices. The measure of “TOT effect” based on the NIA framework is characteristically different from the standard barter TOT concept in many ways. First, indices of NBTOT and its variants are based on the relative commodity price of inter-sectoral sale and purchase by a sector. The ratio of prices received and prices paid, when expressed in indexed form evaluates the deviation in relative prices from base year’s level which is fixed at 100. On the other hand, measures of IRF and “TOT effect” within the NIA framework attempt to quantify the flow of real balance of resource (or income) from a sector due to changes in prices received and prices paid by a sector. The transfer of real balance over a period of time is seen in comparison to the base year, which in this case is set at zero, implying no transfer of resource or TOT gains have occurred due to changes in domestic price levels.

Second, a NBTOT index considers only the bilateral trade between two competing sectors and excludes any intra-sectoral trade within the two. Therefore, TOT improvement for one sector necessarily implies deterioration in the other. In contrast, the NIA framework evaluates TOT position for a sector from its own level in the base year. The measure provided in equation (7.1) can be conceived as hypothetically denoting the TOT situation of a sector vis-à-vis the economy. The TOT effect derived in this fashion can therefore indicate a simultaneous deterioration (or improvement) in both agriculture and industry, if the sectoral TOT has actually moved favouring (against) other component sectors of the economy.

Third, the measure of “TOT effect” focuses on the price of production rather than the price of exchange used in the regular NBTOT measure. The price of exchange is usually approximated by the market price (unit value) of different commodities. On

Table 1: Terms of Trade Effects, Gain (Production) and Gain (Income) in Major Sectors (Base 1980-81)

Sector -> I Agriculture Allied II Industry (WB Definition) III Services (WB Definition) Production Gains Income Gains Production Gains Income Gains Production Gains Income Gains

Years (Rs Crore) As Per (Rs Crore) As Per (Rs Crore) As Per (Rs Crore) As Per (Rs Crore) As Per (Rs Crore) As Per Cent of Cent of Cent of Cent of Cent of Cent of Sectoral Sectoral Sectoral Sectoral Sectoral Sectoral

Real GDP Real GDP Real GDP Real GDP Real GDP Real GDP

1950-51 174.9 0.7 -391.9 -5.7 217.0 1.8
1955-56 -3203.4 -11.7 160.1 1.7 3043.4 20.8
1960-61 -3169.8 -9.9 -4067.2 -12.7 430.7 3.4 25.5 0.2 2739.1 15.0 2083.5 11.4
1965-66 562.5 1.8 960.7 3.1 -1730.7 -9.9 -1532.8 -8.7 1168.2 5.0 1476.1 6.3
1970-71 633.5 1.6 -50.7 -0.1 -1560.9 -7.3 -1892.8 -8.9 927.4 3.2 429.0 1.5
1975-76 -2150.1 -4.8 -4571.3 -10.2 -112.1 -0.4 -1530.8 -6.1 2262.3 6.4 124.4 0.4
1980-81 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
1985-86 -2504.1 -4.6 -1476.4 -2.7 848.0 2.0 1723.9 4.0 1656.1 2.8 2863.9 4.8
1990-91 91.5 0.1 2035.8 3.1 -1498.5 -2.4 340.9 0.5 1407.0 1.7 3900.2 4.7
1991-92 2763.7 4.3 5118.0 8.0 -3223.1 -5.1 -1123.6 -1.8 459.5 0.5 3538.0 4.1
1992-93 873.1 1.3 3255.8 4.8 -1824.2 -2.8 377.2 0.6 951.0 1.0 4145.1 4.5
1993-94 2160.1 3.1 5331.0 7.6 -2787.7 -4.1 79.7 0.1 627.5 0.6 4929.3 5.1
1994-95 4460.7 6.1 9401.5 12.8 -3717.6 -5.0 719.6 1.0 -743.1 -0.7 5748.4 5.6
1995-96 4063.1 5.7 8456.9 11.8 -2877.1 -3.3 1969.0 2.3 -1186.1 -1.0 5544.4 4.7
1996-97 2540.7 3.3 6931.4 8.9 -3238.0 -3.5 1640.4 1.8 697.4 0.5 7699.0 6.1
Sectors -> II Industry (UN Definition) III Services (UN Definition)
Production Gains Income Gains Production Gains Income Gains
(Rs Crore) As Per Cent of (Rs Crore) As Per Cent of (Rs Crore) As Per Cent of (Rs Crore) As Per Cent of
Years Sectoral Real GDP Sectoral Real GDP Sectoral Real GDP Sectoral Real GDP
1950-51 -126.2 -2.4 -48.7 -0.4 -
1955-56 170.9 2.4 3032.6 18.0 - -
1960-61 447.6 4.7 135.8 1.4 2722.2 12.8 1973.1 9.2
1965-66 -1236.1 -9.4 -1086.6 -8.3 673.6 2.4 1029.9 3.7
1970-71 -1367.5 -8.7 -1608.7 -10.2 734.1 2.1 144.9 0.4
1975-76 71.4 0.4 -998.3 -5.3 2078.7 5.0 -408.1 -1.0
1980-81 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
1985-86 -817.4 -2.5 -178.9 -0.5 3321.5 4.8 4766.8 6.9
1990-91 -4228.5 -8.6 -2902.5 -5.9 4137.0 4.2 7143.6 7.3
1991-92 -5325.0 -11.2 -3838.2 -8.1 2561.3 2.5 6252.6 6.1
1992-93 -4582.9 -9.3 -3032.8 -6.1 3709.8 3.5 7555.1 7.0
1993-94 -5682.6 -11.0 -3674.5 -7.1 3522.4 3.1 8683.6 7.6
1994-95 -6757.3 -12.1 -3647.9 -6.5 2296.5 1.9 10115.8 8.3
1995-96 -7267.0 -10.8 -3792.8 -5.6 3203.8 2.3 11306.1 8.3
1996-97 -9120.9 -12.7 -5678.6 -7.9 6580.2 4.5 15017.9 10.2

Notes: (a) The data source is National Accounts Statistics. (b) TOT gain during the base year assumes a situation of no loss-no gain. (c) TOT gain (production) and (income) have been calculated according to Measures I and II, respectively. (d) In the World Bank definition, industry sector comprises of mining, manufacturing; construction; and electricity, gas and water supply and the service sector includes wholesale and retail trade, resturants and hotels; transport, storage and communications; finance, insurance, real estate and business services, and community, social and personal services. In the United Nations definition, construction and electricity, gas and water supply are included in services instead of industry. (e) The indirect taxes and subsidised have not been considered in this table.

the contrary, prices in the NIA framework, particularly the one that represents price received by a sector, is based on the implicit price deflator (IPD) of the sector. These IPDs, which are derived by dividing the sectoral value added at current prices by that at constant prices, signify the price development of the entire sector.

Fourth, there are dissimilarities between the two frameworks as to how the quantum or value of tradeable is to be incorporated for the purpose of developing an income TOT type of measure. The common income TOT index as provided in equation (p 1714), includes the actual volume of exports by a sector. The NIA approach on the other hand provides a hypothetical assessment. That is, it includes the total sectoral real output on the assumption that the sector purchases goods and services in proportion to its value added.

Finally, despite these differences we can see the similarity between the measures of “TOT effect” as per NIA approach and NIRF as provided by Ishikawa (1967a). We have seen earlier that real NIRF (equation 6.1) can be decomposed in two parts, one is the visible NIRF and the other being invisible NIRF caused by changes in inter-sectoral TOT. We observe that the expression of the TOT effect is equivalent to the invisible part of the IRF measure. The distinction if any, is only with regard to how the price variables as well as volume of tradeable have been incorporated.

IV Sectoral Definition and Data Source

Sectoral Definition

The two different concepts of TOT effect, as given in equations

7.1 and 7.2, are formulated using a three-sector classification of the economy consisting of (1) agriculture and allied activities,

(2) industry, and (3) service sectors. To construct this, alternate sectoral definitions based on the World Bank and the United Nations classifications are used, as shown in Table A. As can be seen, according to the United Nations classification system, construction and electricity, gas and water supply are considered part of the services sector and not of industry.

Data Source: All-India

Information concerning the sectoral contributions to the total GDP, at current and constant (1980-81) prices are provided in various issues of the National Accounts Statistics (NAS) brought out by Central Statistical Organisation (CSO). The IPD for sector j (Pva.j) is calculated as ratio of sectoral GDP at current to that of constant prices, both at factor cost. Similarly, the GDP price deflator (P) is worked out as the ratio of economy’s GDP at

vacurrent to that at constant prices, again at factor cost. Our price index for the overall final demand (PD) is a weighted index of price of private final consumption expenditure (PFCE) and the price of total gross domestic capital formation (GDCF). For this, IPDs for PFCE and GDCF are worked out by using the CSO data. The weight used for GDCF (wGDCF) is its share in the GDP at market prices at constant prices, and (1 – wGDCF) is the weight for PFCE.

State Level Data

The state level data on a comparable basis are available for the period 1970-71 and 1995-96. Further, since the GDCF data are not available for all the states on a comparable basis, our state level examination of TOT is confined to analysing the TOT gain (production) measure only. The required data for this purpose are compiled from the official estimates of state domestic product (SDP) as prepared by the respective state statistical bureau or directorates of the various state governments. The SDP estimates for the period 1970-71 to 1980-81 (base: 1970-71) have been assembled from the CSO publication Estimates of State Domestic Product [GoI 1985]. The estimates of SDP for the period 1980-81 to 1995-96 (base: 1980-81) are collected from the office of the CSO in the electronic format as furnished to the CSO by the state statistical agencies. The sector specific IPD in different states are worked out as in the all-India analysis. Further, since it is not possible to chain the two series with different base years, our estimates of the TOT effect refer to two different base periods. That is, TOT estimates from the 1970-71 to 1980-81 refer to the 1970-71 base, whereas estimates from 1980-8l to 1995-96 have 1980-81 as the base year.

Quality of IPD as Price Indices

The measure of TOT effect is based on the NIA principles, which requires the use of the IPD as price indices. Thus, it becomes essential to keep in mind the possible shortcomings of IPD as indicators of price signals in the economy. First, a sectoral IPD truly reflects the price of production provided a double deflation method is used to calculate the real value added of the sector. However, it has not been possible to apply the double deflation method consistently for all the economic sectors in India. This is due to the non-availability or limitations in required data base that constrained the separation and subsequent identification of price and quantity data for all items of inputs and output in certain sectors, particularly in the service segment [GoI 1980, 1989]. In this context, Roychoudhury and Mukherjee (1984) claim that IPDs for the primary and secondary sectors are reasonably representative and realistic as compared to the IPD for the service sector.

Second, the inherent weakness in the IPD for various service sectors persist due to the fact that output in these sectors is typically not well defined in India. That is, while estimating the real value added in these hard-to-measure service sectors, it is generally assumed that output increases proportionately with inputs, thus ruling out productivity growth in these sectors by assumption. This may have led to an upward bias in the IPDs of the service producing sectors.

Table A: Alternate Sectoral Definitions

Three-Sector World Bank Definition United Nations Definition Classification

Agriculture and (1) Agriculture, forestry and (1) Agriculture, forestry and allied activities logging and fishing logging and fishing Industry (2) mining and quarrying (2) mining and quarrying

  • (3) manufacturing (3) manufacturing
  • (4) electricity, gas and water supply
  • (5) construction Services (6) trade, hotel and restaurants (4) electricity, gas water supply

  • (7) transport, storage and com (5) construction
  • (8) finance, insurance, (6) trade, hotel and restaurant real estate and business (7) transport,storageand com services
  • (9) community, social and (8) finance, insurance, real
  • personal services estate and business services

    (9) community, social and personal services

    Economic and Political Weekly April 29, 2006

    Third, the methodology for estimating different value added aggregates has been revised several times by the CSO. While the revised estimates for previous years are generally provided along with each revision of the NAS data, the extent of such revisions is considerable at the sectoral level [GoI 1989]. It has been claimed in this context that the changes in concepts, database and methodology in NAS do not pose a serious problem for longterm analysis over the period 1950-51 to 1996-97 with 1980-81 as base [Sivasubramonian 2000].

    It is apparent that similar caveats also apply to the sectoral IPDs at the state level, which are worked out from the CSO’s state domestic product data. In spite of these limitations, IPDs have been used by Ahluwalia (1979,1985), Roychoudhury and Mukherjee (1984) to analyse the relative price behaviour, and recently by Misra and Hazell (1996), Misra (1998) and Acharya (2001) to examine the agricultural TOT at the level of all-India and individual states. However, it may be kept in mind that if the IPDs are used in the calculations, the derived TOT index will capture the relative sectoral price variability from a producer perspective, unlike the wholesale or consumer prices which reflect the perspective of consumer income and welfare.

    V Movements in TOT Effects

    All-India Analysis

    Patterns in TOT effects: Table 1 provides the estimates of TOT effect for the agriculture, industry and service sectors of the Indian economy. It appears that the service sector in India has by and large experienced favourable TOT effects (i e, TOT gains) over the entire study period. The extent of the favourable TOT effect

    Table 2: Time Trend in TOT Effects (1961-76 and 1982-97)

    Sectors TOT Gain (Production) TOT Gain (Income) 1961-76 1982-97 1961-76 1982-97

    Agriculture and allied 0.81(2.43)* 0.73(8.41)* 0.75(2.08)* 1.12(9.37)* Industry (WB) -0.56(-2.04)* -0.40(-6.61)* -0.56(-2.29)* -0.04(-0.51) Services (WB) -1.03(-5.63)* -0.18(-4.65)* -1.02(-6.11)* 0.21(6.07)* Industry (UN) -0.68(-2.19)* -0.88(-11.76)* -0.68(-2.45)* -0.55(-5.52)* Services (UN) -0.92(-4.77)* 0.02(0.33) -0.91(-5.19)* 0.41(8.44)*

    Notes: (1) The coefficient of time trend (t) is given in the table.

  • (2) The time trends are obtained by regressing the TOT effects (TOT gains as percentage of real output) on time in linear form.
  • (3) * Indicates significant t values at 10 per cent level of significance.
  • for services is more pronounced when we define the sector according to the UN classification system. Concomitantly, the adversity of industrial TOT is greater when electricity, gas and water supply and construction subsectors are excluded to define the aggregate industry sector.

    Domestic TOT at the beginning of the 1950s remained favourable to agriculture as well as the services sector, and unfavourable to the industry sector. A distinct change occurred during the mid1950s whereby the TOT effect became negative for agriculture along with a coexisting improvement for the industry sector. This pattern continued till the mid-1960s, whereby some considerable TOT gain was also noticed for the services sector in this period. The second major turnaround in the pattern of Indian domestic TOT effect occurred during the mid-1960s, which was marked by a distinct improvement for agriculture along with a deterioration for industry and a decline in the extent of TOT gains in services. The agricultural sector continued to experience positive TOT effects till about mid-1970s, after which there was a slump in the agricultural TOT effect which carried on till the beginning of the 1990s. During the 1990s, sectoral TOT effects again turned favourable to agriculture and distinctively negative to industry. The TOT effects on services remained positive, though it fell considerably in the 1990s. A graphical plot of TOT gain (production), as per measure I, is provided in Figure 1. Trend of TOT effects: We analyse the statistical trend of TOT effects in these three broad sectors during two sub-periods.5 It may be noted that the earlier studies have detected a pattern for the agricultural NBTOT during the period between early or mid-1960s and mid-1970s. Similarly, the recent NBTOT estimates of the government agencies (the CACP and DES) and Thamarajakshi (2000) have revealed an improvement of agricultural NBTOT between early-1980s and mid-1990s. We therefore examine the trend of TOT effects referring to the sub-periods 1960-61 to 1975-76 and 1981-82 to 1996-97. Table 2 provides this information.

    We notice a positive trend in the TOT effects for agriculture and allied sector during both the sub-periods, i e, the TOT gain (production) and TOT gain (income) measure indicated statistically significant upward movements in this sector. On the contrary, the trend in both the measures of TOT effects for the aggregate industry sector is downward movements in each of the two sub-periods. This result is robust to alternate sectoral definitions. The TOT effect towards services reflected a positive trend during the 1980s after having moved downward between

    Table 3: Trend of Sectoral Terms of Trade Effects in Major States, 1972-81 and 1982-91

    Agriculture and Allied Industry Services 1971-72 to 1980-81 1981-82 to 1990-91 1971-72 to 1980-81 1981-82 to 1990-91 1971-72 to 1980-81 1981-82 to 1990-91

    Kerala 0.44 Gujarat 1.92 Orissa 3.69 Orissa 2.16 Uttar Pradesh 3.49 Punjab 0.73 Maharashtra 0.02 Bihar 1.38 Rajasthan 2.98 Himachal Pradesh 1.38 Bihar 2.09 Haryana 0.48 Madhya Pradesh -0.02 Maharashtra 0.97 Madhya Pradesh 2.74 Andhra Pradesh 1.17 Karnataka 1.41 West Bengal 0.24 Himachal Pradesh -0.09 Karnataka 0.69 West Bengal 2.12 Kerala 0.97 Andhra Pradesh 1.25 Himachal Pradesh 0.22 Rajasthan -0.49 Uttar Pradesh 0.31 Punjab 1.76 Tamil Nadu 0.46 Tamil Nadu 1.24 Kerala 0.20 Haryana -0.54 Rajasthan 0.30 Haryana 1.39 West Bengal 0.21 Orissa 1.08 Maharashtra 0.07 Punjab -1.24 Madhya Pradesh 0.28 Andhra Pradesh 1.32 Uttar Pradesh 0.16 Punjab 1.03 Tamil Nadu -0.22 Orissa -1.26 Andhra Pradesh -0.13 Gujarat 1.21 Punjab -0.05 West Bengal 0.91 Rajasthan -0.45 Bihar -1.26 Haryana -0.16 Karnataka 0.97 Rajasthan -0.09 Haryana 0.54 Orissa -0.46 Andhra Pradesh -1.28 Tamil Nadu -0.39 Bihar 0.88 Madhya Pradesh -0.15 Gujarat 0.52 Karnataka -0.55 Karnataka -1.32 Punjab -0.41 Uttar Pradesh 0.84 Karnataka -0.43 Maharashtra 0.48 Andhra Pradesh -0.61 Gujarat -1.56 Orissa -0.48 Kerala 0.52 Haryana -0.63 Himachal Pradesh 0.20 Bihar -0.74 Tamil Nadu -1.67 West Bengal -0.58 Tamil Nadu 0.04 Maharashtra -0.79 Rajasthan -0.69 Uttar Pradesh -0.76 West Bengal -1.84 Himachal Pradesh -0.92 Himachal Pradesh 0.03 Gujarat -0.84 Kerala -0.79 Gujarat -0.77 Uttar Pradesh -1.94 Kerala -0.92 Maharashtra -0.49 Bihar -1.27 Madhya Pradesh -1.62 Madhya Pradesh -0.82

    Notes: (1) Trends are derived by regressing TOT gains (production) on time in linear form. (2) States rearranged by ranking the magnitude of trend in a descending order. (3) Industry and service sectors are defined according to the World Bank definition.

    Figure 1: TOT Effects in Agriculture and Allied, Industry and Services: 1950-51 to 1996-97, (Base: 1980-81)

    TOT Gain as Per Cent of Real Sectoral GDP

    30 25 20 15 10 5 0 –5 –10 –15

    950/51 1958/59 1966/67 1974/75

    Year

    Agr & Allied

    Industry

    Note: Sectoral Classification as per World Bank Definition.

    1960-61 to 1975-76. In particular, the TOT gain (income) measure indicates the presence of statistically significant positive trend for both the World Bank and UN definition.

    State Level Analysis

    Patterns in TOT effects: The results on sectoral TOT effect for 15 state economies of India during specific years in the 1970s, 1980s and 1990s have been provided in the Appendix Table A1. The TOT effect on agriculture has been negative during the 1970s, which seems to have continued till about the mid-1980s in most of the states. The exceptions are Haryana, Rajasthan and to some extent Madhya Pradesh and Maharashtra, where some positive agricultural TOT effect is noticed during the 1970s. Beginning 1990-91, a favourable shift in agricultural TOT effect is seen in some of the states. The states of Bihar, Gujarat, Maharashtra, Rajasthan and West Bengal have a distinctively positive agricultural TOT effect since early-1990s, which may mark the beginning of policy reforms in agriculture. Some improvements in agricultural TOT effect during the post-reform phase are also seen in the states of Haryana, Himachal Pradesh, Karnataka, Kerala and Orissa. Conversely, a negative TOT effect on agriculture is noticed in Andhra Pradesh, Madhya Pradesh, Punjab, Tamil Nadu and Uttar Pradesh. The TOT situation of the aggregate agricultural and allied activities sector in these states is more or less similar, excepting Himachal Pradesh and Orissa, where the allied activities (viz, forestry and logging, fishing) had positive TOT effects during the 1990s, despite a negative agricultural TOT effect.

    A generally negative TOT effect towards manufacturing is observed during the 1980s, which persisted till the mid-1990s in most states, with the exceptions of Kerala, Tamil Nadu and particularly Andhra Pradesh. The TOT effects of the aggregate industrial sector in different states mostly represent a pattern similar to the one depicted by their respective manufacturing sectors. During the mid-1990s too, a generally negative industrial TOT effect is seen in most states, barring Andhra Pradesh, Kerala, Tamil Nadu,

    Table 4: Pattern of Sectoral TOT Effects across Indian States

    Sectors Statistically Significant Trend of TOT Effect Upward Trend (Favourable TOT) Downward Trend (Unfavourable TOT)

    Agriculture Bihar, Gujarat, Haryana, Himachal Pradesh, Karnataka, Maharashtra, Punjab, Rajasthan, Uttar Pradesh and All-India. None

    Agriculture and Allied Andhra Pradesh, Bihar,Gujarat, Haryana, Karnataka, Kerala, Maharashtra, Orissa, Punjab, Rajasthan, Uttar Pradesh, West Bengal and All-India None

    Manufacturing Andhra Pradesh, Kerala Bihar, Gujarat, Haryana, Himachal Pradesh, Karnataka, Maharashtra, Madhya Pradesh, Punjab, Rajasthan, Uttar Pradesh, West Bengal and All-India.

    Industry (WB definition) Andhra Pradesh, Uttar Pradesh Bihar, Gujarat, Haryana, Karnataka, Maharashtra, Punjab, Rajasthan, West Bengal and All-India Services (WB definition) Punjab, West Bengal Bihar, Gujarat, Karnataka, Kerala, Madhya Pradesh, Orissa, Uttar Pradesh and All-India

    Industry (UN definition) Andhra Pradesh, Kerala Bihar, Gujarat, Haryana, Himachal Pradesh, Karnataka, Madhya Pradesh, Maharashtra, Punjab, Rajasthan, Uttar Pradesh, West Bengal and All-India

    Services (UN definition) Himachal Pradesh, Maharashtra, Punjab and All-India Andhra Pradesh, Bihar, Gujarat, Karnataka, Kerala, Rajasthan, Tamil Nadu and Uttar Pradesh

    Note: Based on statistically significant trends during 1981-82 to 1995-96. Source: Derived from Tables 2, 3 and Appendix Table A-3.

    Economic and Political Weekly April 29, 2006 Orissa and Uttar Pradesh. The sectoral TOT effect of the aggregate service sector was positive in as many as nine states during the 1970s. This pattern continued in Madhya Pradesh, Punjab, Rajasthan, Tamil Nadu and West Bengal during the 1990s but in other states, a negative service TOT effect is noticed. Trend of TOT effects: The statistical trends of sectoral TOT effect are provided in Table 3 for each state during the 1970s and 1980s. We have mentioned earlier that construction of a single TOT series for the whole study period was not possible at the state level. Therefore, the trends have been calculated separately using the decadal TOT series between 1971-72 and 1980-81 and between 1981-82 and 1990-91. Further, the states have been ranked in descending order by the growth rates of TOT effect. The table shows that sectoral TOT effects during the 1970s has registered a sustained downward movement for agriculture and the allied sector and upward movement for both industry and service sectors in most of the states. A noticeable turnaround in the trend direction of TOT movements occurred at the state level during the 1980s, whereby the sectoral TOT began to move with a positive trend in agriculture along with a coexisting declining trend in both industry and service sectors.

    We further examine the statistical trend by including the estimates of sectoral TOT effects for the post-reform years, i e, by calculating the trend in state-level sectoral TOT series during the period between 1981-82 and 1995-96. The results are provided in Appendix Table A2. We noticed a statistically significant positive trend in agricultural TOT effect for nine states. On the other hand, a significant downward movement for the TOT effects on manufacturing was seen for 11 out of 15 states in our sample. The only exceptions are Andhra Pradesh and Kerala. There is evidence of a significantly declining TOT effect for the aggregate industry sector in eight states, and similarly, a significant downtrend in the TOT effect of services sector was observed for seven states. A sustained improvement for service TOT effect is seen only in the states of Punjab and West Bengal. When the aggregate industry sector is defined according to UN classification (by excluding EGW and construction sectors), more states see a statistically significant downturn in the industrial TOT effect. Concomitantly, there are fewer states that reflect a statistically significant declining trend in service TOT effect. The information on the statistically significant time trend in sectoral TOT effects of different state economies is classified in Table 4.

    VI Comparison with NBTOT Series

    Earlier analyses of domestic TOT in India have been undertaken on the basis of examining the NBTOT between agriculture and industry (non-agriculture) sector, viz, Thamarajakshi (1969, 1994), Tyagi (1987), Mungekar (1992), Palanivel (1999) and also by two government agencies, viz, the CACP and DES. These NBTOT series after connecting them with a common base have been reported in Deb (2002a). We observe that the decade between mid-1960s and mid-1970s has been marked with a favourable TOT shift for agriculture in all the NBTOT estimates. This is true for the present estimates of “TOT effect” on agriculture as well. A subsequent turnaround of TOT against agriculture has likewise been noticed in most of the NBTOT estimates and in the present estimates of agricultural “TOT effect”. Finally, the results also reflect a favourable turnaround in agricultural TOT during the post-reforms period in India, similar to the series provided by Thamarajakshi (2000) and the two government agencies, CACP and DES. As far as the issue of agricultural TOT at the state level is concerned, only a few studies are available on a comparable basis across the Indian states. Recently, Acharya (2001) provided analogous estimates of agricultural TOT encompassing 17 major states by using the IPD of agriculture and nonagriculture from the SDP (CSO) data. Similar to our results, his analysis also reveals that the compound growth rate of agricultural TOT recorded a positive growth in almost all the states during the period 1980-81 to 1997-98.

    VII Conclusions

    This paper can broadly be seen as an attempt to examine the implications of incorporating the service sector into the analysis of domestic TOT in India. Thus, we have provided an analysis of domestic TOT movements by formulating a three-sector analysis consisting of agriculture, industry and service sectors of the economy. It was felt that it is important to study domestic TOT in a three-sector setting, in view of the structural transformation of the Indian economy signified by the expansion of the service segment over the last five decades. The estimates of “TOT effect” accruing to agriculture, industry and service sectors of the economy are provided at the all-India level and also at the level of 15 major states in India. The time period of analysis is 1951-97 at the all-India level and 1971-96 at the level of individual states. Our all-India results indicate that domestic TOT effect in recent years has been positive for agriculture and services sector and negative for the industry sector. The TOT effects for the industry sector have not been positive since the mid-1960s. The state level estimates indicate certain regional variations in the pattern of domestic TOT movements in India. A statistically significant improvement in agricultural TOT during 1980-81 to 1995-96 has been observed for nine out of 15 states. Concomitantly, a significant deterioration in the manufacturing TOT effect was seen for 11 states during the same period except in the states of Andhra Pradesh and Kerala. We found evidence of a declining TOT effect on aggregate industry sector in eight states. Similarly, a significant decline in services TOT was observed for seven states.

    Overall, the results derived by incorporating the service sector into the TOT analysis suggest that sector can potentially play an important role in the determination of domestic trade terms. In fact, our all-India analysis suggests that domestic TOT gains have consistently accrued in the service sector in almost the entire study period. It appears that if one wants to know what is going on in the domestic TOT front in India, it is time that one starts taking the service segment seriously. One also infers that the practice of estimating TOT in the two-sector approach may remain insufficient to capture the impacts of relative price change in economies that have a large service component. Towards this end, we suggest that the present accounting framework can be extended to re-examine the TOT effects in other economies where the question of accumulation, growth and distribution associated with the domestic TOT change remained important.

    Finally, it has been observed that the examination of a number of economic concerns has been informed by the analysis of domestic TOT in the context of developing economies. For instance, a large number of studies in India have attempted to analyse the extent to which TOT changes interact with various policy indicators of the economy, viz, the interaction of agriculture-industry

    Appendix Table A1: Terms of Trade Effect in Major States of India (Base: 1970-71 and 1980-81)

    Agriculture Agriculture and Allied Manufacturing Industry Services Gain in Gain as Per Cent Gain in Gain as Per Cent Gain in Gain as Per Cent Gain in Gain as Per Cent Gain in Gain as Per Cent Lakh of Sectoral Lakh of Sectoral Lakh of Sectoral Lakh of Sectoral Lakh of Sectoral Rupees Real GDP Rupees Real GDP Rupees Real GDP Rupees Real GDP Rupees Real GDP

    Andhra Pradesh
    1970-71 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
    1975-76 -17729.4 -11.3 -16050.3 -10.0 6315.0 21.9 8669.3 19.4 7381.0 8.0
    1980-81 -12714.4 -8.0 -10998.3 -6.8 3436.5 9.4 4368.7 7.5 6629.6 5.4
    1980-81 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
    1985-86 -47413.1 -13.0 -35538.6 -9.4 494.8 0.4 24118.6 14.4 11419.9 3.2
    1990-91 -27606.8 -6.3 -17308.6 -3.8 24315.2 17.9 29222.8 12.8 -11914.2 -2.5
    1995-96 -29216.8 -5.8 -16983.0 -3.2 28674.2 19.4 27667.5 10.8 -10684.4 -1.8
    Bihar
    1970-71 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
    1975-76 -6680.8 -5.0 -5775.8 -4.2 7619.9 29.5 7440.1 12.2 -1664.2 -3.0
    1980-81 -165.1 -0.1 378.9 0.3 -3363.9 -9.8 -4656.6 -6.3 4277.7 5.8
    1980-81 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
    1985-86 -8882.7 -2.4 -8018.2 -2.1 -2206.8 -1.7 6572.0 3.1 1446.2 0.6
    1990-91 34310.6 8.7 30847.3 7.2 -14000.5 -8.4 -17760.2 -6.1 -13087.1 -4.2
    1995-96 61938.6 15.8 55495.9 12.9 -27514.5 -17.2 -45254.7 -14.9 -10241.3 -2.8
    Gujarat
    1970-71 0.0 0.0 0.0 0.0 10.0 0.0 0.0 0.0 0.0 0.0
    1975-76 -9077.8 -8.9 -8283.0 -7.9 3636.6 8.5 4696.3 8.0 3586.7 4.4
    1980-81 -9251.2 -8.3 -7770.1 -6.8 6291.4 10.8 5387.6 6.7 2382.5 2.2
    1980-81 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
    1985-86 -16544.0 -7.3 -10424.6 -4.2 -23069.2 -10.4 5122.0 1.8 5302.6 1.8
    1990-91 22689.6 8.4 29500.2 10.0 -51149.5 -16.4 -9334.4 -2.4 -20165.8 -4.9
    1995-96 58568.7 22.3 64665.6 22.2 -93142.7 -19.4 -57733.4 -10.0 -6932.2 -1.3
    Haryana
    1970-71 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
    1975-76 1335.5 2.2 1388.1 2.2 -143.6 -1.2 -1169.3 -6.4 -218.8 -0.8
    1980-81 1094.7 1.6 1449.1 2.1 1223.1 6.6 -292.9 -1.1 -1156.2 -3.0
    1980-81 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
    1985-86 -13232.0 -6.5 -12869.9 -6.3 -52.0 -0.1 5444.9 5.7 7424.9 6.2
    1990-91 -5940.0 -2.3 -6527.8 -2.5 -4299.3 -3.8 3505.0 2.6 3022.8 1.7
    1995-96 -14856.2 -5.4 -15874.4 -5.7 3907.5 2.8 19216.1 11.6 -3341.6 -1.5
    Himachal Pradesh
    1970-71 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
    1975-76 -931.7 -6.5 -461.5 -2.9 253.1 19.0 361.2 7.7 100.3 1.3
    1980-81 -1590.5 -12.0 -551.2 -3.7 81.8 6.4 -70.0 -1.4 621.2 7.2
    1980-81 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
    1985-86 -2567.5 -8.3 -412.0 -1.1 -563.7 -9.5 519.7 2.8 -107.8 -0.4
    1990-91 -5284.9 -13.9 -3905.8 -8.5 -1154.5 -9.9 2917.6 10.7 988.1 2.4
    1995-96 2365.4 5.9 2803.9 5.5 -6095.6 -35.4 -4149.9 -10.5 1346.0 2.6
    Karnataka
    1970-71 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
    1975-76 -9116.7 -8.4 -8849.3 -7.9 1727.5 4.9 988.8 1.8 7860.5 15.8
    1980-81 -7085.1 -6.5 -5643.4 -5.0 -1101.0 -2.1 1334.8 1.8 4308.6 6.7
    1980-81 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
    1985-86 -28428.3 -11.0 -12133.3 -4.5 2872.2 2.8 15508.1 10.5 -3374.8 -1.3
    1990-91 -542.8 -0.2 16593.1 5.3 -20451.9 -12.5 -2579.4 -1.2 -14013.7 -3.7
    1995-96 7698.4 2.0 27153.8 6.7 -35097.0 -17.8 -6656.5 -2.6 -20497.3 -4.1
    Kerala
    1970-71 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
    1975-76 -4413.1 -7.0 -2387.4 -3.6 580.7 3.2 352.3 1.4 2035.1 4.0
    1980-81 -1226.2 -2.0 2317.7 3.7 1483.2 5.9 1897.6 5.5 -4215.2 -6.5
    1980-81 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
    1985-86 -10034.0 -7.3 -8516.6 -5.7 -4232.2 -7.2 8740.9 9.1 -224.3 -0.1
    1990-91 -24040.8 -13.6 -16283.3 -8.7 2862.8 3.6 12863.5 10.1 3419.8 1.6
    1995-96 6182.0 3.1 25514.2 11.8 343.2 0.3 4382.7 2.3 -29896.8 -9.2
    Madhya Pradesh
    1970-71 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
    1975-76 -236.6 -0.2 -111.4 -0.1 1917.0 8.9 2823.4 6.5 -2712.0 -4.6
    1980-81 3141.6 2.6 5208.2 4.0 1498.6 5.6 6315.6 11.8 -11523.7 -15.2
    1980-81 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
    1985-86 -53096.9 -13.7 -48565.4 -11.4 502.3 0.6 39686.2 25.7 8879.2 3.7
    1990-91 -54552.8 -11.4 -31375.6 -6.2 -20885.0 -11.8 21183.9 7.9 10191.7 3.0
    1995-96 -48351.7 -9.3 -36216.6 -6.7 -22726.7 -11.1 30619.0 9.0 5597.5 1.3
    Maharashtra
    1970-71 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
    1975-76 -690.7 -0.5 1292.1 0.9 -4404.0 -3.5 -2284.5 -1.4 992.4 0.6
    1980-81 167.3 0.1 3391.3 2.0 -8629.4 -4.9 -5091.1 -2.4 1699.9 0.8
    1980-81 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
    1985-86 7037.6 1.9 15461.0 3.7 -32092.8 -6.3 6542.5 1.0 -22003.5 -2.8
    1990-91 1757.8 0.3 20860.7 3.4 -91908.3 -12.0 -37456.2 -3.9 16595.5 1.4
    1995-96 87016.9 13.5 100969.5 14.4 -149956.8 -13.2 -26188.9 -1.9 -74780.6 -4.2
    (Contd)
    1720 Economic and Political Weekly April 29, 2006
    Appendix Table A1: Terms of Trade Effect in Major States of India (Base: 1970-71 and 1980-81) (Contd)

    Agriculture Agriculture and Allied Manufacturing Industry Services Gain in Gain as Per Cent Gain in Gain as Per Cent Gain in Gain as Per Cent Gain in Gain as Per Cent Gain in Gain as Per Cent Lakh of Sectoral Lakh of Sectoral Lakh of Sectoral Lakh of Sectoral Lakh of Sectoral Rupees Real GDP Rupees Real GDP Rupees Real GDP Rupees Real GDP Rupees Real GDP

    Orissa 1970-71 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1975-76 -2451.5 -3.5 -2314.9 -3.2 1208.8 16.0 2544.3 18.6 -229.5 -0.9 1980-81 -7605.2 -9.0 -5872.5 -6.7 3096.5 23.6 6000.5 32.2 -128.0 -0.4 1980-81 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1985-86 -1791.3 -0.9 1507.4 0.7 -2361.3 -6.1 3402.9 5.1 -4910.3 -3.8 1990-91 -6742.3 -4.5 3752.9 2.2 -1447.9 -3.0 9465.5 9.4 -13218.4 -8.1 1995-96 17848.0 10.2 23578.1 11.7 -7343.5 -11.0 -6932.0 -5.1 -16646.1 -7.5 Punjab1970-71 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1975-76 -6072.5 -6.3 -6042.6 -6.2 2454.7 13.9 4641.4 15.9 1401.3 2.8 1980-81 -11095.2 -9.9 -11263.9 -10.0 4063.6 15.5 7913.5 19.6 3350.5 4.4 1980-81 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1985-86 -32434.4 -10.9 -32102.5 -10.6 -4331.2 -5.2 2202.5 1.9 29899.9 17.5 1990-91 -25969.7 -7.3 -27747.7 -7.7 -11236.9 -9.2 -1127.2 -0.7 28875.0 13.2 1995-96 -7932.4 -1.9 -10198.5 -2.4 -45765.4 -24.6 -42579.6 -17.4 52778.0 20.5 Rajasthan1970-71 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1975-76 734.3 0.8 930.5 1.0 508.1 3.5 414.3 1.5 -1344.8 -2.7 1980-81 19.4 0.0 379.2 0.4 2011.9 12.6 4366.5 13.5 -4745.7 -8.0 1980-81 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1985-86 -9859.3 -4.0 -9153.4 -3.7 -2977.0 -4.7 2949.9 2.7 6203.5 3.9 1990-91 -3566.9 -0.9 -2810.5 -0.7 72.4 0.1 1686.3 1.0 1124.3 0.4 1995-96 21166.8 5.4 20037.0 4.9 -17407.2 -15.2 -25795.1 -11.2 5758.1 1.6 Tamil Nadu 1970-71 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1975-76 -11485.9 -11.6 -12082.2 -11.8 6542.6 12.5 9330.2 13.0 2752.0 2.9 1980-81 -16219.6 -21.2 -16426.5 -20.7 2476.0 3.3 2619.3 2.8 13807.3 12.1 1980-81 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1985-86 -11110.0 -4.9 -10496.5 -4.5 -6139.4 -2.5 7312.6 2.4 3183.9 0.8 1990-91 -37792.9 -13.9 -38542.7 -13.6 36823.7 12.3 24549.5 5.9 13993.2 2.5 1995-96 3591.0 1.0 15518.6 4.3 -19355.4 -6.2 -13491.1 -3.0 -2027.4 -0.3 Uttar Pradesh 1970-71 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1975-76 -21335.7 -8.1 -21344.8 -7.9 2215.5 5.2 4646.1 6.5 16698.7 13.9 1980-81 -24385.2 -8.0 -23706.9 -7.7 5538.7 8.9 -1031.1 -0.9 24738.0 16.5 1980-81 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1985-86 -46042.4 -5.8 -46530.5 -5.8 -18630.8 -8.5 11779.6 3.8 34750.9 6.0 1990-91 -13337.0 -1.4 -13906.7 -1.4 -23830.6 -7.0 20967.7 4.5 -7061.0 -0.8 1995-96 -22035.6 -2.1 -24374.5 -2.3 -40247.1 -10.6 28816.1 5.7 -4441.5 -0.5 West Bengal1970-71 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1975-76 -17379.1 -11.1 -16824.8 -10.2 13175.6 22.8 16386.8 19.8 438.0 0.4 1980-81 -30781.3 -17.7 -28922.3 -15.9 14331.9 21.4 20966.9 22.4 7955.4 5.9 1980-81 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1985-86 149.7 0.0 11457.7 3.0 -20115.0 -9.3 3458.2 1.1 -14915.9 -3.2 1990-91 17229.8 4.4 26562.7 5.9 -1352.3 -0.5 -13753.1 -3.4 -12809.5 -2.1 1995-96 39014.6 7.3 66756.4 11.1 -41828.1 -12.8 -76376.4 -14.7 9620.0 1.1

    Notes: (1) TOT gain in the base year assumes a situation of no loss or gain. (2) TOT gains are calculated as per Measure I. (3) Estimates prior (subsequent) to 1981refer to 1971 (1981) as base. (4) The industry sector comprises mining, manufacturing, electricity, gas and water supply and construction. The service sector includes transport, storage and communication; trade, hotels and resturants; banking and insurance; real estate, ownership of dwelling and business services;public administration and other services.

    Appendix Table A2: Time Trend of Sectoral Terms of Trade Effects in Major States, 1982-92 to 1995-96

    States Agriculture Agriculture and Allied Manufacturing Industry (WB) Services (WB) Industry (UN) Services (UN)

    Andhra Pradesh 0.10 0.20 **** 1.67 * 0.43 *** -0.53 ** 1.49 * -0.66 * Bihar 2.17 * 1.87 * -1.66 * -1.61 * -0.69 * -1.70 * -0.85 * Gujarat 2.15 * 2.08 * -0.87 * -1.07 * -0.36 ** -1.17 * -0.40 * Haryana 0.39 *** 0.29 **** -0.94 ** -0.62 **** -0.14 -0.93 ** -0.06 Himachal Pradesh 0.67 *** 0.30 -1.78 * -0.40 0.06 -1.47 * 0.21 **** Karnataka 0.83 * 0.87 * -1.43 * -0.53 ** -0.58 * -1.29 * -0.29 * Kerala 0.11 0.75 ** 0.71 * 0.06 -0.56 ** 0.73 * -0.66 ** Madhya Pradesh 0.11 0.19 -1.47 * -0.54 -0.39 ** -1.30 * 0.02 Maharashtra 0.11 * 1.08 * -1.09 * -0.58 * -0.04 -1.06 * 0.16 *** Orissa -0.02 0.31 **** -0.04 0.41 -0.36 * -0.38 0.03 Punjab 0.37 ** 0.27 **** -1.33 * -1.29 * 0.79 * -1.31 * 0.49 ** Rajasthan 0.59 * 0.54 * -0.79 ** -1.03 * -0.11 -0.94 ** -0.37 *** Tamil Nadu 0.22 0.41 0.18 -0.14 -0.14 0.30 -0.39 * Uttar Pradesh 0.45 * 0.40 * -0.45 ** 0.20 ** -0.73 * -0.47 ** -0.34 ** West Bengal 0.35 0.56 *** -0.62 *** -1.25 * 0.41 * -0.55 **** -0.17

    Notes: (1) Trends are obtained by regressing TOT gains (production) on time in linear form.

    (2) *, **, ***, and **** indicate significant t-values at 1 per cent, 5 per cent, 10 per cent and 20 per cent respectively.

    growth; rural wage and poverty; technology (HYV) adoption, further analysis on these issues. The comparable set of agriculspread of irrigation, private investment, government expenditure, tural TOT estimates for the state economies are very useful in and total factor productivity growth in agriculture, etc. Towards modelling TOT in a cross-sectional framework. The cross-section this effect, the TOT results of this study can be extended for analysis is particularly important if one takes into account the fact that most of the econometric work associated with domestic TOT analysis of policy issues has been performed using time series TOT data at the all-India level. With state level estimates one can test certain policy hypotheses that are best modelled using a cross-sectional framework. For instance, the implications of TOT change can be examined in relation to the rural-to-urban migration process and changes in the workforce compositions or incidence of rural and urban poverty – all of these being concerns that can be studied only using a data set which is available for select points of time.

    mr'

    Email: postsurajit@yahoo.co.in

    Notes

    [This paper is based on a chapter of my PhD thesis. I have benefited fromthe inputs provided by K L Krishna, Amit Bhaduri, T C A Anant, UmaRoychoudhury, S Tendulkar and V Pandit. The usual disclaimers apply.]

    1 According to data published by the ADB (2002), the primary sector’s sharein GDP fell from 44.5 per cent to 24.3 per cent over the period 1970 to2001. The sectoral share of industry and service sector rose from 23.9per cent to 26.8 per cent and 31.6 per cent to 49.0 per cent, respectivelyduring this period.

    2 This aspect is discussed in Deb (2002a).3 The derivation is as follows. Recall that the expression of real NIRF, inthe case of an import excess balance was given by:

    ME 1 11

    Real NIRF = −= R + E − E (since R + E = M)

    PPP P P

    MEMM E

    11 PE 1 =

    R + . E − E

    PPP P

    M EM E

    11 ⎛ PE ⎞

    = R + .E ⎜⎜ −1⎟⎟ PM PE ⎝ PM ⎠

    4 We recall that the expression of real NIRF in the case of an import excess balance was given as:

    11 ⎛ P ⎞ Real NIRF = R + E ⎜⎜ E −1⎟⎟ (6.1)

    P PP

    ME ⎝ M ⎠

    With the first part of the right hand side representing the visible part andthe other implying the invisible part, which is due to changes in intersectoral TOT (compare equations (I) and (I’) in Ishikawa 1967, p 297).On the other hand, the formula for TOT gains as per the NIA framework(equation 7.1) can also be derived as follows:

    1⎛ Pva. j ⎞ Sectoral TOT Gains = [X va ′ . j ( va. j − Pva )] = X va ′ . j ⎜⎜ −⎟⎟

    P 1

    P P

    va ⎝ va ⎠

    X va. j ⎛ Pva. j ⎞ = P ⎜⎜ P −1⎟⎟ (7.1.1) va. j ⎝ va ⎠

    It can be observed from equation (6.1) and (7.1.1) that the measures of“TOT effect” and invisible NIRF have similar expressions. That is, while,P and P are analogous to P and P, representing the prices received

    va.jvaEM

    (or price of exports) and prices paid (or price of imports), the exportproceeds (or marketable surplus) of a sector is equivalent to the total valueadded in the sector.

    5 The TOT effects revealed absence of any statistically significant trendduring the overall study period. This is consistent with the earlier assertionsmade by Thamarajakshi (1994) and Palanivel (1999) that NBTOT in Indiadid not bear any distinct trend in the long run. See Deb (2002a) for details.

    References

    Acharya, S S (2001): ‘Terms of Trade for the Agricultural Sector in Indiaat the State Level’ in S S Acharya and D P Chaudhri (eds), Indian Agricultural Policy at the Cross Roads, Rawat Publishers, New Delhi.

    Ahluwalia, I J (1979): Behaviour of Prices and Outputs in India, Macmillan.

    – (1985): ‘Interrelationships between Agriculture and Industry’ in Indian Economic Association (1985), Conference Volume of the 68th. AnnualConference held in Ahmedabad.

    Asian Development Bank (2002): Asian Development Outlook, Oxford University Press Inc, New York.

    Bjerke, K (1968): ‘Some Reflections on the Terms of Trade’, Review of Income and Wealth, 14 (2).

    – (1972): ‘Some Reflections on the Intersectoral Terms of Trade for Denmarkin the Period 1949-65’, Review of Income and Wealth, 18 (4).

    Deb, Surajit (2002a): ‘The Debate on Agriculture-Industry Terms of Tradein India’, Working Paper No 109, Centre for Development Economics,Delhi School of Economics, University of Delhi, Delhi.

    Derksen, J B D (1980): ‘Measuring International and Intersectoral Terms ofTrade: Some Methodological Issue’, Review of Income and Wealth, 26 (3). Dorrance, D (1950): ‘Income Terms of Trade’, Review of Economic Studies, 16 (1).Government of India (1980): National Accounts Statistics: Sources and Methods, Central Statistical Organisation, Ministry of Planning, New Delhi.

  • (1985):Estimates of State Domestic Product,Central Statistical Organisation,Ministry of Planning, New Delhi.
  • (1989):National Accounts Statistics: Sources and Methods, Central Statistical Organisation, Ministry of Planning, New Delhi.
  • Ishikawa, S (1967a): Economic Development in Asian Perspective, Kato Bumeisha Printing, Tokyo.
  • (1967b): ‘Resource Flow between Agriculture and Industry: The ChineseExperience’, The Developing Economies, 5 (1).
  • Kahlon, A S and D S Tyagi (1980): ‘Intersectoral Terms of Trade’, Economic and Political Weekly, 15 (52).
  • (1983): Agricultural Price Policy in India, Allied Publishers, New Delhi. Karshenas (1995): Industrialisation and Agricultural Surplus, Oxford University Press, New York.Kindleberger, C P (1956): The Terms of Trade: A European Case Study, John Willy, New York.Lee, T H (1971): Intersectoral Capital Flows in the Economic Developmentof Taiwan: 1895 to 1960, Itacha, New York. Misra, V N (1998): ‘Economic Reforms, Terms of Trade, Aggregate Supply andPrivate Investment in Agriculture’,Economic and Political Weekly, 33 (31).
  • Misra, V N and P B R Hazell (1996): ‘Terms of Trade, Rural Poverty,Technology and Investment: The Indian Experience, 1952-53 to 1990-91’,Economic and Political Weekly, 31(13), Review of Agriculture.

    Mody, A (1979): ‘Resource Flow between Agriculture and Non-agriculture:Critique of an Estimate’, Indian Journal of Agricultural Economics, 34 (4).

    – (1981): ‘Resource Flow between Agriculture and Non-agriculture in India,1950-70’, Economic and Political Weekly, 16 (10, 11 and 12).

    Mody, A, S Mundle and K N Raj (1985): ‘Resource Flows from Agriculture:Japan and India’ in K Ohkawa and G Ranis (eds), Japan and theDeveloping Countries, Basil Blackwell, Oxford.

    Mundle, S (1977a): ‘Intersectoral Resource Flow in Post-colonial India’,Indian Economic Review, 12 (2).

    –(1977b):‘Home Market, Capitalism in Agriculture and the Drain of AgriculturalSurplus’, Economic and Political Weekly, 12 (26).

  • (1981): Surplus Flows and Growth Imbalances: The Intersectoral Flowof Real Resource in India, 1951-1971, Allied Publishers, New Delhi. Mundle, S and K Ohkawa (1979): ‘Agricultural Surplus Flow in Japan, 18881937’, The Developing Economies, 17 (3).Mungekar, B L (1992): The Political Economy of Terms of Trade, HimalayaPublishing House, Mumbai.
  • Olgaard, A (1966): Growth Productivity and Relative Prices, North Holland, Amsterdam.
  • (1981): ‘Concepts and Problems of Consistency in Terms of Trade Analysis’,Review of Income and Wealth, 27 (2).
  • Palanivel, T (1992): Inter-sectoral Terms of Trade and Resource Flows: Some Implications for Agricultural Development, Income and Poverty in India, PhD thesis, Department of Economics, University of Delhi, Delhi.

    – (1999): ‘The Long-Run Trends in the Terms of Trade between Agricultureand Non-Agriculture in India: Some Alternative Methods’, Indian Journal of Economics, 69 (315).

    Rasmussen, P N (1957): Studies in Intersectoral Relations, North Holland Publishing, Amsterdam.

    Roychoudhury, U D and M Mukherjee (1984): National Accounts Information System, Macmillan.

    Sheng, Y (1992): Intersectoral Resource Flow and China’s Economic Development, Macmillan.

    Sivasubramonian, S (2000): The National Income of India in the Twentieth Century, Oxford University Press, New Delhi.

    Staehle (1951): ‘Some Notes on the Terms of Trade’, International Social Science Bulletin (UNESCO), 3 (1).

    Stuvel, G (1956): ‘A New Approach to the Measurement of Terms of TradeEffects’, Review of Economics and Statistics, 38 (3).

    Thamarajakshi, R (1969): ‘Intersectoral Terms of Trade and Marketed Surplusof Agricultural Produce, 1950-51 to 1965-66’, Economic and Political Weekly, 4 (26).

  • (1994): Intersectoral Relationships in a Developing Economy, Academic Foundation, Delhi.
  • (2000): ‘National Agricultural Policy: Confusion on Ends and Means’, Economic and Political Weekly, 35 (35 and 36).
  • Tyagi, D S (1987): ‘Domestic Terms of Trade and Their Effect on Supply andDemand of Agricultural Sector’, Economic and Political Weekly, 22 (13).

    Economic and Political Weekly April 29, 2006

    Dear Reader,

    To continue reading, become a subscriber.

    Explore our attractive subscription offers.

    Click here

    Back to Top