An Empirical Investigation of Twin Deficits Hypothesis: Evidence from India

The term “twin deficits” was conceived in the 1980 s in the United States (US) when both the CAD and the budget deficit increased significantly. As a result of this co-movement, a significant portion of the deterioration in the external balance was attributed to the emergence of huge budget deficits. This causal relationship is known as the TDH (Salvatore 2006).

The current account surpluses were found in the years 1973, 1976, 1977, 2001, 2002 and 2003. Here, the annual year in India refers to the financial year, which starts from 1st April (month of the current year) to 31st March (month of the next year). For e.g., the year 1973 refers to 1973–1974. This pattern is followed throughout the papers.

The fiscal deficit of the Central government has increased from 3.17 per cent of gross domestic product (GDP) in 1970 to nearly 5 per cent of GDP in 2013. Similarly, CAD as a percentage of GDP (CADGD) has increased from nearly 1 per cent in 1970 to more than 5 per cent in 2012. After the year 2012, both fiscal deficit of the Central government as percent of GDP (FSDFG) and CADGD has gradually declined in the recent year 2016, i.e., FSDFG has gone to below four per cent, while CADGD also declined to below one per cent. During the period from 1970 to 1979 (1970s), the averages of FSDFG and CADGD were 3.09 and 0.09 respectively. Then, the averages of FSDFG increased to 7.21 and CADGD ALSO rose to 1.96 during the year from 1980 to 1989 (1980 s). However, the averages of FSDFG decreased from 6.25 in the year 1990–1999 (1990 s) to 5.14 in 2000–2009 (2000s), while the average of CADGD also fell from 1.35 to 0.56 during that time. During 2010–2013, the averages of both FSDFG (5.33) and CADGD (3.60) was increased. However, recently during the period from 2013-2016, the averages of FSDFG was 4.3 per cent, while CADGD was slightly higher than one per cent.

Due to global financial integration and financial openness, the standard economic theory states that the savings of any country would flow to countries with the most productive investment opportunities. Thus, domestic savings rates would be uncorrelated with domestic investment rates. However, Feldstein and Horioka (1980) found that domestic saving rates and domestic investments rates are highly correlated for OECD countries, which has launched a debate regarding the degree of financial integration and financial openness in the industrial as well as developing countries since then. The research findings of Feldstein and Horioka (1980), which are contrary to the economic theory, have to be referred as Feldstein-Horioka puzzle.

It assumes M-X is CAD and G-T is fiscal deficit. Thus, in the next step, M − X is shown as CAD and G − T is represented as fiscal deficit in the Eq. (4). Here, G is total government expenditure, while T refers to total revenues, which includes both Tax and Non-Tax revenues. The difference between the total revenue and expenditure is called as budget deficit. However, in India, fiscal deficit is measured as the excess of total expenditure over total receipts other than borrowings. It is to be noted that the current account balance includes trade balance, and net income and transfer flows. In India, trade balance constitutes a major portion of it, while net income and transfer flows are very negligible.

Exchange rate refers to Indian rupees per unit of the US dollar.

Trade openness is measured as an increase in the size of a country’s traded sectors in relation to total output. Here, it is defined as the ratio of sum of exports and imports to GDP [(Exports + Imports)/GDP].

Fiscal deficit can be of the gross fiscal deficit of the Central government or the combined gross fiscal deficits of the Central and State governments. The study has carried out the analysis separately for both of these deficits along with other control variables. The empirical analysis throughout the paper are done bysing real variables.

The details of these variables are explained in the Sect. 3.1. All of these models are estimated by using real variables.

It has not taken both trade openness and exchange rate simultaneously in model A as these two variables are highly correlated. To avoid multicollinearity, and also to check robustness of our result, it has estimated it separately. As the selected sample periods are small, it is appropriate to take crucial variables in a model rather than taking all variables at a time.

Model A and B are estimated for the period from 1970–1971 to 2013-14. However, models C and D are estimated using data from 1980–1981 to 2013–2014, due to data constraint. Because the data for the combined fiscal deficits of the Central and State governments are available from 1980–2081 onwards [Source: Table No. 113, Hand Book of Statistics (HBS), Reserve Bank of India (RBI)].

Due to annual observation, the maximum lag is selected as 2. Thus, the maximum value for ‘m’ is 2.

The error correction term can be obtained as: ECM = \( Y_{t} - (a_{0} + \sum\nolimits_{i = 1}^{m} {a_{1} Y_{t - i} } + \sum\nolimits_{i = 0}^{m} {a_{2} X_{t - i} } + \sum\nolimits_{i = 0}^{m} {a_{3}^{\prime } Z_{t - i} } ) \).

It estimates the speed at which the system corrects its previous period disequilibrium for converging towards its long run equilibrium. For e.g., if the coefficient of the error correction term is − 0.5 and significant, it implies that the system corrects its previous period disequilibrium at a speed of 50 per cent to reach at the steady state.

Source: Table no. 244, HBS, RBI data base.

Except interest rate hike, capital inflows (needed to bridge the current account gap) can be uncertain in the presence of inflation, large fiscal deficit, country characteristics, political uncertainty etc. (Trachanas and Katrakilidis 2013; Busse and Hefeker 2007; Anoruo and Ramchander 1998) India is not an exception to these circumstances as it faced a high persistence of fiscal deficits, high volatility in inflation in the study period, and uncertainity is bad for India (Bhagat et al. 2016). The study has used long term interest rate on government of India securities, which is more reliable and stable. Therefore, an increase in interest rate could attract foreign direct investment in terms of transfer of knowledge and better technology etc. to enhance productivity in the long run. Thus, CAD would be declined due to the production of import substitution products and rise in exports in the long run.

Source: Table no. 229, HBS, RBI.

During the period 1980–1989, the average of the saving-investment rate gap was − 1.8 per cent, then it was declined to − 1.4 per cent during 1990–1999, and further the average of the gap between saving and investment drastically reduced to -0.6 per cent in 2000–2009. But the gap has jumped to more than three percent during 2010 and 2013, which is attributed due to comparatively more fluctuations in the investment rate than the saving rate.

As the investment rate has also increased from 15.1 per cent to 33.8 per cent during the same period.

An increase in economic growth in the previous period could enhance the production of goods and services, which would compensate the rise in the domestic demand (fall in imports of foreign goods) in the short run causing less CAD. Similarly, higher economic growth would enhance the collection of total revenues, which would reduce fiscal deficits, and can accumulate more foreign exchange reserves in the economy. Thus, in the later period CAD will be reduced.

In the pre-reform period, financial sector of the Indian economy had been affected by complete public ownership, complex set of regulations, and entry barriers in the banking sector. Systematic financial sector reforms have been undertaken under the new economic policy reforms of 1991 in terms of liberalization of the entry and expansion of private sector banks, dismantling of administered interest rates, reduction in mandatory ratios, i.e., cash reserve ratio (CRR) and statutory liquidity ratio (SLR), etc. Especially, since 1993, domestic private sector banks, branches of foreign banks, and foreign portfolio investment have been permitted. A series of reforms were undertaken in the capital markets like the securities and exchange board of India (SEBI) was given statutory power in 1992, the national stock exchange (NSE) was established in 1994, the national securities clearing corporation (NSCC) was established in 1995 etc. All these scenarios have smoothen and developed the financial openness of the economy (Bhavani and Bhanumurthy 2012, pp: 5–8).

Lagrange multiplier test is used for residual Serial correlation; Ramsey's RESET test using the square of the fitted values represents model misspecifications; Normality test is based on a test of skewness and kurtosis of residuals; Hetero-scedasticity test is based on the regression of squared residuals on squared fitted values.