R.S. Tiwari

Pressure of Domestic Demand and Export Performance in India

Homepage Manual Contents Navigation

METHODOLOGICAL ISSUES

Estimated stock of a commodity, in this case bicycles, with domestic consumers is assumed as a proxy variable for domestic demand. In fact, real proxy for domestic demand of individual consumers is the proportionate fraction of their total income to be effectively spent on bicycles' purchasing. However, due to scarcity of adequate data the present method, also used by NCAER in estimating the demand for scooters, has been followed.¹⁵ For the purpose of our estimation, we assume that life time of bicycles prevailing in Indian market must be fifteen years.

The estimate of domestic stocks with consumers has been accumulatively worked out as total output+imports-exports over the period 1960-61 to 1970-71. Total output of bicycles which is readily available at current producers prices has been converted at 1961 base.¹⁶ Import and export values of bicycles originally available at CIF and FOB prices, are found to be objectively uncomparable. And due to lack of appropriate conversion factors these data have been converted at 1961 wholesale prices for all the years under study.¹⁷ Moreover, personal disposable income originally available at current prices has been purposefully converted at a comparable base by using comparable implicit price deflator over the period 1960-61 to 1970-71.

Similarly, relative price-indices of bicycles have been worked out by percentage ratios of bicycles' wholesale price indices to that of general wholesale price indices at a comparable base. Estimated stocks of consumers are treated as a dependent variable whereas, personal disposable income and relative prices are explicitly used as independent explanatory variables. The regression equation in the compact form would be:

(1. Equation in values)
y = alpha + bX + CZ
y = -2518.96 - 6.62 X + 0.03 AZ
(3.604) (0.004)*
R² = 0.97* ..................

(2. Equation in quantities)
^y = alpha + bx + CZ
^y = -23281.47 - 65.21 X + 0.25 Z
(34.40)*** (0.037)*
R² = 0.97* ..................

where,
y = domestic demand for bicycles (consumers' stocks) in million rupees at 1961 = 100
^y = domestic demand for bicycles (consumers' stocks) in quantity 1000' nos.
X = relative price indices of bicycles in rupees at 1961 = 100
Z = personal disposable income in million rupees at 1961 = 100
* indicates significant at 1 per cent level
*** indicates significant at 10 per cent level
Figures in parenthesis denote the standard errors of estimate.

To reinforce our regression results regarding the sign of explanatory variables compound growth rates over period 1960-61 to 1970-71 at 1961 prices have been calculated and are given below.

Regression results and annual compound growth rates explain that only increase in the level of personal disposable income would have stimulated domestic demand of bicycles. However, the ß coefficient of relative prices for explaining domestic demand of bicycles are found to be statistically insignificant in terms of values and inversely associated in terms of quantities respectively.

The domestic demand for bicycles estimated from equations (1) and (2) are used as an independent explanatory variable for explaining the behaviour of export performance over the period 1960-61 to 1974-75 in the subsequent analysis. Export of bicycles is explained by domestic demand (consumers' stocks) estimated from equations (1) and (2) indices of per capita GDP and imports at 1961 prices. Indices of per capita GDP at 1961 market prices have also been considered as other independent explanatory variable to reflect the purchasing power ability for selective import-countries of bicycles from India, viz., Iran, Afghanistan, Nigeria, the USA, Netherlands, Singapore, Sri Lanka and Indonesia.¹⁸ Similarly, indices of imports for major import-markets have also been introduced at a comparable base as another independent variable to represent their internal demand patterns. Moreover, in order to observe the impact of Government policy on export performance dummy variable would be incorporated in some regression models. Standard multiple regression model is used to test the hypothesis in this analysis.

Information regarding exports and imports (volumes and values) have been collected from "Monthly Statistics of Foreign Trade", Department of Commercial Intelligence, Calcutta, Government of India.²⁰ Total output data for bicycles over the period 1961 to 1969, excepting 1967, have been taken from "ASI Census Sector", furnished from CSO, New Delhi.²¹ Total output data for remaining years, viz., 1967. 1970 and 1971, have been projected on the basis of annual average compound growth rates. Additionally, information regarding personal disposable income and prices have been selected from "RBI Bulletin"²² and "Wholesale Price Indices"²⁸ respectively. However, data regarding personal disposable income for ending two years i.e., 1974 and 1975, have been projected by the same method of annual compound growth rates. Information relating to imports and import prices and mid-year population for selective import countries are carefully collected from IFS (International Financial Statistics)²⁴ and "Monthly Bulletin of Stattstics"²⁵ furnished by IMF and United Nations respectively. Moreover, the data for GDP at current prices and its implicit deflators for selective import markets have been collected from "World Table".²⁶

Assumptions

1. Bicycle in home as well as in export markets is assumed to be homogenous in character.
2. Main objective of domestic manufacturers is assumed to be profit maximisation.

The Analysis

Export of bicycles in India has widely been observed to be determined by sets of controllable and non-controllable variables. Factors substantially affecting export of Indian bicycles from demand side are thought to be explained largely by sets of international forces, viz., relative prices, per capita GDP anti internal demand patterns of selective import-markets. Whereas, the factors on the supply side are widely determined by sets of internal variables, viz., subsidy-cash, open, disguised, nominal and explicit-premiums, direct tax concessions and import entitlement and replenishment schemes. In fact, the combined effects of both-supply bottlenecks and demand constraints-would reasonably explain the determination of export performance.

Since data for bicycle prices in importing counlries are not found to be adequately available, the influence of relative prices on export of Indian bicycles has been objectively ruled out. Moreover, the quantification of policy impact on the quantum of exports product is practically impossible, mainly because of its fluctuating nature.²⁷ Nevertheless, dummy variable would be incorporated in some regression models in order to capture the policy influences. Multiple regression models are being given here as follows to test the proposed hypothesis.

ß Coefficients of estimated domestic demand of bicycles (i.e., the stock of domestic consumers) are found to be statistically significant in equations 3, 5, 7 and 9. In addition, ß coefficients of indices of pattern of internal demand for selective import markets in this case bicycles are marked to be statistically significant in cases 3,4,5 and 6. Whereas, ß coefflcients of indices of their per capita GDP are statistically reliable only in two cases, i.e., equations 5 and 6. However, coefficients of policy variable are seemed to be statistically insignificant and therefore, unreliable throughout the period under study. R², in all the cases are significant, shows the percentage of variations explained in this subsequent analysis. Durbin Weston Statistics (DW), straight forwardly, reflects the absence of auto correiation for most of the data used in various regression models.

Thus, pressure of internal demand operating both in export and import markets determines the export performance in terms of values, whereas, the behaviour of internal demand prevailing only in selective import markets explains substantially the export performance in terms of quantities. In fact, during the reference period, export demand for Indian bicycles in these import markets might have been stimulated partly due to the availability of export of bicycles from India at a compar- atively cheaper price compared to the rest of the world export markets and partly due to the lack of production of bicycles in their local import markets. Further, the effect of increased percapita GDP in their selective import markets has, however, outweighed the favourable eflects of domestic demand prevailing in both the markets of export and import regions in explaining the rate of export performance of Indian bicycles.

The prospect of rate of export performance of bicycles in its developed and developing import markets might have been adversely affected due to increased level of industrialisation in the former and due to observed inflationary trend as a result of continuous spurt in oil prices in the latter.

MODELS
Coefficients of determination of export, performance of bicycles in India with respect to domestic demand, policy variable, indices of per capita GDP of selective markets and indices of their internal demand patterns during the period of 1960-61 to 1974-75 at 1961 prices

Model I: Simple Linear Regressions

Missing table. Please write me an email, if you are interrested in this table.

Model II: Logerthemic Regressions

Missing table. Please write me an email, if you are interrested in this table.

Model V: Share Model with Policy Variable

Missing table. Please write me an email, if you are interrested in this table.

The upshot of our analysis, i.e., performance of relative share of export of bicycles is being explained dominantly by the pressure of internal demand operating in Indian economy. Other explanatory variables are found to be statistically insignificant in this particular context. Annual average compound growth rates are also presented to supplement our regression results.

Compound growth rates of indices of GDP and home demand as compared to estimated domestic demand seem to be relatively ineffective in determination of compound growth rates of exports.

Hence, domestic demand in association with internal demand pattern seems to be a dominant explanatory variable in explaining the export of bicycles as compared to the other structural determinants from import markets and Government Policy. Variations of domestic demand might have led to tne expansion of domestic production through the improvement in productivity. Production coefficients, given below, explain that input requirements of coal and anodes to produce one physical unit of bicycles, excepting 1965, have declined in 1969 as compared to 1961. Which explains in other words, that an increase in productivity would have led to an increase in bicycles' production.

Forgoing analysis, with the help of multiple regression model, compound growth rates and productivity trends explains that variation of domestic demand has been positively associated with export performance. And, therefore, Linder's hypothesis has once again been reaffirmed in the case of export performance of bicycles in India at 1960-61 prices.