Testing of Savings-Growth Relationship in India: An Application of Cointegration and Error Correction Techniques

Article excerpt

This paper mainly attempts to examine the causal nexus between Gross Domestic Product (GDP) and savings (at 1993-94 prices) in India, using the annual observation from 1950-51 to 2002-03. Although a number of studies have shed light on this issue using time series data, the present study attempts to analyze the relationship between the variables using cointegration and causality models. The results show the presence of cointegration between the GDP and savings series implying the presence of a stable long-run relationship between them. Having verified that GDP and savings are cointegrated, the Granger test indicates the absence of any causal relationship between the variables in log-difference form. Despite the absence of Granger causality, a negative coefficient of the error correction term in the regression equation of savings on GDP is observed. This negative coefficient signifies that savings converge to its long-run equilibrium level.

(ProQuest: ... denotes formulae omitted.)

Introduction

There has been a growing interest among the planners and policy makers on the role played by savings and investment in the economic development of both developed and developing countries. Models based on theories of endogenous growth pioneered by Roemer (1986) and Lucas (1988) predict that higher savings and investment rates can permanently raise growth rates. More recent research (e.g., Levine and Renelt, 1992; De Long and Summers, 1993; Easterly and Rebelo, 1993; and King and Levine, 1994), finds that the investment rate is one of the most important determinants of economic growth. Attanasio et al. (2000) examine the dynamic relationship between economic growth, investment and savings rate, using annual time series for a large cross-section of countries. Employing a variety of samples and econometric techniques, they consistently find that growth Granger causes savings, although the effect appears to be quantitatively weak. They also find that an increase in savings rate does not always precede increase in growth (Loayza et al., 2000, p. 401).

Schmidt-Hebbel et al. (1996) review several theories on the direction of causation between savings and growth, ranging from the classical permanent-income and life-cycle hypotheses to the more recent and less conventional models. The newer models that emphasize slow changing consumption habits (Carroll et al., 1995), a mixture of strong consumption habits with uncertain incomes, valuing both consumption and wealth (Cole et al., 1992; Fershtman and Weiss, 1993; and Zou, 1993a and 1993b), broadly suggest that growth drives savings.

Rodrik (2000) finds strong evidence that growth precedes savings in the pooled annual data for the countries with saving transitions. The evidence using a five-year average is somewhat weaker. As for the reverse relationship, the result indicates, if anything, a negative, perverse effect from savings to growth. Growth Granger causes savings, while savings (negatively) Granger cause growth (Rodrik, 2000, p. 495).

Lahiri (1989) finds that the rate of growth of personal disposable income is a significant determinant of private savings in all the countries in his sample of Asian countries, including India. Lahiri bases his empirical finding on individual time series analysis, for each country in his sample. He claims that such an approach has an advantage over a panel-based analysis in that the marginal response of the savings rate to various factors need not be assumed uniform across countries. Muhleisen (1996) uses a Vector Autoregression (VAR) process in logarithms to jointly model the relationship between the private savings rate and growth and between private and public savings (i.e., Ricardian equivalence). Using Granger causality tests, Muhleisen shows that growth leads to both higher public and private savings rates.

Causation is two-way if savings also increase growth. Growth leads to more savings, which in turn enhance growth. …