AN EMPIRICAL STUDY ON THE SOCIAL
SECTOR DEVELOPMENT OF TAMIL NADU
Dr. Debesh Bhowmik
Retired Principal And Associated with The Indian Economic Association, and
The Indian Econometric Society, P.O.-Pritinagar, Road No-1, District-Nadia,
W.B.,741247, Mobile No-07602072569,
[email protected], [email protected]
ABSTRACT
In this paper, the author attempted to relate social sector expenditure with net state domestic
product and per capita net state domestic product of Tamil Nadu from 1990-91 to 2016-17
with the help of econometric models of semi-log regression model, Bai-Perron structural
model, Hodrick-Prescott-Filter Model, Johansen cointegration test and Vector Error
Correction Model taking RBI data. The paper concludes that social sector expenditure of Tamil
Nadu has been catapulting at the rate of 12.61% per year during 1990-91-2016-17. This social
sector expenditure contains four significant upward structural breaks in 1996,2003,2007 and
2011, respectively. Hodrick-Prescott Filter model minimized its cyclical patterns where
frequency response function showed one peak only. On the other hand, net state domestic
product and per capita net state domestic product have been stepping up at the rate of
13.78% and 11.67% per year during 1990-91-2016-17. The ARIMA (1, 1, 1) model of social
sector expenditure of Tamil Nadu showed non-stationary, and its projection for 2031 proved
steady progress. Granger causality test proved that there is unidirectional causality running
from net state domestic product to social sector expenditure and from social sector
expenditure to per capita net state domestic product of Tamil Nadu during 1990-91-2016-17.
Johansen cointegration test confirmed one cointegrating equation among social sector
expenditure, net state domestic product, and per capita net state domestic product in the
specified period. The cointegrating equation is significant and tends towards an equilibrium
where the coefficient of error correction is significant at 5% level, and the speed of
adjustment is shown as 8.43% per annum. But the vector error correction model is unstable,
non-stationary,non-normal, and serially correlated. It is verified that there is a long run
association among social sector expenditure, net state domestic product, and per capita net
state domestic product of Tamil Nadu during the survey period. Wald test from the system
equation verified that there is short-run causality running from social sector expenditure to
net state domestic product and per capita net state domestic product and vice versa.
Keywords- Social Sector Expenditure & Growth, GDP, GDP per capita, Co-integration, Granger
Causality, Vector error correction
JEL Classification-E32,H5,H62,O4,I0
Economic Challenger// ISSN 0975-1351/ Issue 84, July-Sept. 2019
The Social Sector Development of Tamil Nadu
3