Uncertainty,Trade and Growth: An Indian Experience

Abstract

Over the years, emerging economies have extensively followed a liberal trade regime and witnessed unprecedented economic growth. At the same time, global uncertainty has also reached to the pinnacle and has gripped almost every country within its ambit. Trade openness though seems to be channelising negative spillover effects of global uncertainty, yet continues to flourish. Therefore, the economic implications of uncertainty in presence of unprecedented trade openness remain a moot question which lays the basis of this study. Using autoregressive distributive lag (ARDL) approach, the empirical estimates suggest that macroeconomic effects of uncertainty are effectively mitigated by trade openness.

JEL Codes: C22, F14, C22

Keywords

Trade openness global uncertainty economic growth ARDL

I. Introduction

Today the global economy is characterised by increasing volatility, economic uncertainty and lacks clarity of future activities. Over past few decades, uncertainty in global economy has risen considerably. This is partially driven by the increasing trade negotiations and proposals among investors, politicians and market participants. The increasing integration not only contributes significantly but also has exposed its participants to the global uncertainty. The effects of domestic or bilateral issues are felt globally. For instance, the recent incidents like Sino-US standoff and outbreak of Corona virus have sent shock waves throughout the world. In particular, these shocks have been largely felt in countries which are fairly connected to the global economic system through trade linkages. Earlier, the advent of financial crisis and great recession in 2008 left serious global repercussions which took several years to recover. The developing world stands more vulnerable given its weak institutional infrastructure. The switch from boom to bust in developing countries is increasingly determined by the uncertainty emanating from industrial world. This, at times, triggers a strong economic downturn in considerably open second and third world countries.

Given this, the basic objective of extracting positive gains from liberal trade regime stands at odds. On one hand, liberal trade is considered sine qua non for growth and continues to flourish by reaching to the new dimensions of modern economic societies. Countries are increasingly pursuing their social improvements, economic ambitions as well as the political animosities through economic integration. The predominant message from bourgeoning literature of trade–growth relationship indicates positive effects of trade (Romer, 1986; Lucas, 1988; Stokey, 1988; Dollar, 1992; Lee, 1995; Coe & Helpman, 1995; Krueger, 1998; Krishna & Mitra, 1998; Kraay, 1999; Froning, 2000; Were, 2015; Sheikh & Malik, 2021). However, on the flipside, the recent experience of global uncertainty shocks indicates that trade openness serves as a medium to transmit these shocks. Globally well integrated countries with significant trade linkages seem to be more vulnerable to external shocks.

Therefore, a thorough analysis is mandated to ascertain the trade–growth relationship under the shadows of global economic uncertainty. In this backdrop, this study attempts to examine the role of trade openness and global uncertainty in determining economic growth in India.

India provides a compelling case to study this aspect in some detail. Firstly, being an emerging country, with huge population and growing external sector, it remains a moot question as to how the impact of global uncertainty is felt in India. Secondly, notwithstanding the high priority for industrial sector, India remains substantially dependent upon the imports, particularly the high technology goods. Thirdly, the level of institutional development is relatively low as compared to developed countries. The fact that Indian economic growth trajectory is shaped by ‘POW’ trinity—political stability, oil prices and world economic prospect makes it highly susceptible to global uncertainty. This invigorates further interest to empirically examine the trade–uncertainty–growth relationship in Indian context.

In the onset it is important to set the limits of this study. The objective of this study is two-fold. First, we investigate the effects of trade openness and global uncertainty on economic growth in India. Then we proceed to expand our analysis to examine whether trade openness increases or reduces the vulnerability of Indian economy to the external shocks. We contribute to the literature in three different ways. First, we contribute to the thin literature on the impact of global uncertainty on economic growth. Most studies of this genre have focused upon the occasional crisis and have followed piece-wise approach (Cavallo & Frankel, 2008; Hooper & Kohlhagen, 1978). Secondly, this paper belongs to the mushrooming literature examining the role of trade openness in mediating the effects of external shocks on economic growth. Finally, the study belongs to a bourgeoning literature which explores the determinants of economic growth and the role played by trade openness therein.

Our results suggest that, in the long run, uncertainty and trade openness insignificantly affect the economic growth. While, in the short run, uncertainty drives business cycles and trade openness has a transitionary growth effect. Further, trade openness mitigates the negative consequences of global uncertainty in the long run as well as short run.

The rest of the article is structured as follows: Section 2 reviews the existing literature; Section 3 introduces the global uncertainty and the proxies available; Section 4 presents the empirical methodology adopted to analyse the impact of trade openness and uncertainty on growth. Section 5 presents the analysis and interpretation of empirical results. Section 6 summarises and presents the concluding remarks.

II. Review of Literature

Based on the impact of global uncertainty, existing literature is divided and provides number of alternate explanations. While the generalisation that globalisation leads to crisis appeals to many yet many raise strong objections to such generalisation. Skeptics provide different justifications by arguing that heightened uncertainties may weaken country’s export market that may trigger downward trend in its economic output. In one of the early contributions to this field, Friedman (1968) argues that uncertainty in monetary policy has a negative impact on economic growth. Potentially economic growth can be affected via increase/decrease in both demand for exports and the import prices. Clark (1973) postulated that uncertainty especially with regard to exchange rate may render the exporters unable to forecast the foreign price, thereby can fluctuate foreign earnings in an unpredictable fashion. Though hedging may contain the uncertainty to certain extent, yet it is impossible to eliminate it completely. Many argue that loss in trade credit specifically for imports results in trade shrinkages. Therefore, uncertainty is likely to reduce the flow of goods and services and thereby reduces economic growth. Hooper and Kohlhagen (1978), while analysing the impact of uncertainty on price and volume of bilateral and multilateral trade of various industrial countries also reached the similar conclusion. Condon et al. (1984) undertook a case study of post reform crisis in Chilean economy and analysed it through multisector general equilibrium model. Comparing the model generated growth path with that of the actual, the study states that due to rising import prices, Chile witnessed successive external shocks particularly oil price shocks. The dwindling export prices further exacerbated the shocks and led to the severe crisis in Chilean economy. Rodrick (1999) argued that trade openness may lead to political and social conflicts in developing countries by making them vulnerable to external shocks. Being institutionally weak, these countries may not be able to cope with high frequency global shocks. The ‘real options’ theory also predicts that heightened uncertainties cause delays in irreversible decisions involving adjustment costs. Also, the precautionary saving curve of households shift outwards and consequently reduces their consumption in uncertain situation (Carroll, 1997; Kimball, 1990).

Contrary to this, the proponents of free trade openness argue that trade in fact reduces the vulnerability to external uncertainty and higher export to GDP ratio enables better accommodation of external shocks. Cavallo and Frankel (2008) examined the link between external vulnerabilities and trade openness. Using instrumented variable for trade openness to expunge endogeneity, the study establishes causal link running from closed economic structure to financial instabilities. On empirical examination, the study finds trade openness acting as an absorbent of the external shocks. Moreover, strong trade linkages reduce the probability to default and encourage international investors not to pull out. Rose (2005) offers empirical evidence of how strongly trade linkages between countries act as deterrent against sovereign defaults. The study concludes that threatened penalty of downgrading trade relations defines the reluctance of sovereign to default. Countries with high interdependence on each other are less likely to default as compared to the countries which are mutually independent. Moreover, trade openness reduces the trade cost by removing the barriers and improving competitiveness. In a general equilibrium model, Martin and Rey (2006) illustrate that if trade costs are high for emerging economies, financial liberalisation may expose them to financial crises and demand volatility. In such a situation, trade liberalisation reduces the trade costs which not only stabilises the demand but also reduces the vulnerability to financial crises. Fujita (2007) reformulated international trade model based on real option theory and examined how foreign exchange rate uncertainty affects growth of exporting country. The study postulates that both growth rate and fluctuations in the welfare of an exporting country will increase with the incremental exchange rate uncertainty.

Many studies argue that the impact of external shocks differ across countries depending upon country characteristics. To this, Sachs (1985) differentiated between Asian and Latin American experience of external debt crises of early 1980s. The study suggested that the crises were less distortionary to Asian economies than their Latin American counterparts. This was primarily due to higher ratios of exports to output held by Asian economies which enabled them to absorb the external shocks better. A variant of this idea is also provided by Balassa (1981) who examined 28 developing economies which faced negative external shocks during the period of 1974–1978. Differentiating between the outward oriented and closed economies, Balassa posit that outward oriented countries respond differently and tactfully to external shocks than closed economies. While not denying the fact that open countries faced severe shocks more frequently, the study concludes that open countries are better able to cope with such adversities than their closed counterparts. Moreover, following an adverse shock, decline in growth is temporary in open economies as against the permanent decline in closed ones.

However, these studies analysed uncertainty largely in a piecewise and discontinuous manner. These studies have modelled only the occasional but hard hitting crises episodes with effects largely on bilateral trade relations. This is partially because of inherent difficulty involved in measuring uncertainty due to its directly non-observant nature. Since the turn of the century, there has been an expulsion of global events occurring at high frequency which has led to the development of new strand of literature attempting to capture its economic implications. This strand of literature flourished at a faster pace particularly since the seminal work by Bloom (2009) who offered a structural framework to examine growth–uncertainty link. The study modelled uncertainty proxied by stock market volatility (VIX) as second-moment shocks to demand in a vector autoregression (VAR) framework. On the macroeconomic effects of uncertainty, the study suggests that these have a large real impact and generate a substantial drop and rebound in output and employment over next two quarters. Since then several empirical studies have found countercyclical link between uncertainty and economic growth. The noted and often-cited studies of this genre include Baker et al., (2013), Jurado et al. (2013), Leduc and Liu (2012), Bijsterbosch and Guérin (2013) and Caggiano et al. (2013) to name a few. Baker et al. (2013) constructed a new measure of economic policy uncertainty (EPU) based on newspaper references, upcoming expiry of tax provisions and forecasters’ disagreement. Assessing the impact of EPU on investment and employment in firms significantly exposed to it, the study finds EPU is negatively associated with investment and employment. Moreover, the study postulates that the national level innovations in EPU foreshadow sizeable declines in GDP and employment. Leduc and Liu (2012) examined the mechanism that macroeconomic effects of uncertainty follow. To this, Leduc and Liu (2012) argued that increase in uncertainty reduces both employment as well as inflation (INF). These macroeconomic effects occur partly through aggregate-demand channel. Bijsterbosch and Guérin (2013) followed two-step nonlinear approach to investigate the evolution of macroeconomic and financial variables during episodes of high uncertainty. Using regime-switching model to identify the high uncertainty episodes in US, the study finds uncertainty associated with significant decline in economic growth, INF, bond returns and stock prices. In addition to this, uncertainty also increases unemployment. Caggiano et al. (2013) also examined the influence of uncertainty shocks on US unemployment dynamics in post-World War II era. Using non-linear VAR modelling, the study portrays that the effects of uncertainty on employment are negative and much larger when nonlinearities are admitted to play their role. Caggiano et al. (2020) estimated the symmetric effect of US policy uncertainty on Canadian unemployment rate in booms and busts. The study finds that jumps in US policy uncertainty significantly drive up the uncertainty in Canada and affect employment rate negatively. However, the impact is much severe in slack periods than in booms. In line with this, Alessandri and Mumtaz (2014) find uncertainty more fatal during financial distress.

Colombo (2013) analysed the strength of external and internal policy uncertainty in Euro Area. The study proxies the external uncertainty by US EPU measured by Baker et al. (2013) and compares it with Euro Area’s internal policy uncertainty. The study finds external shocks more influential than the internal ones as indicated by Euro Area’s more responsiveness to US EPU than its own. Klobnera and Sakke (2014) also used Baker et al. (2013) index to estimate spillovers of policy uncertainty. For six developed countries, spillovers accounted from one fourth to one half of the uncertainty dynamics. Further, it finds the United States and the United Kingdom as large exporters of shock spillovers. Moore (2016) constructed a broader uncertainty index for Australia using news-based indices, stock market volatility index and disagreement among the earning forecast analysts. Upon weighing its impact on Australian economy, the study finds that both employment and investment slows down while precautionary savings spike up in response to uncertainty shocks. A similar study by Greig et al. (2018) shows that New Zealand economy responds negatively to the uncertainty shocks. The point estimates indicate that output falls from two to four percentage points with every incremental unit of uncertainty which persists up to three to four quarters. In the Indian context, uncertainty and its consequent implications on economic growth and investment were analysed by Bhagat et al. (2013). The point estimates show that if uncertainty reduces to 2005 level, the growth rate and investments would increase by 0.56 and 1.36 percentage points, respectively.

III. Uncertainty and its Measurement

Uncertainty, as Knight (1921) puts it, is people’s inability to make future predictions—is a recent phenomenon. Until recently, there was limited volatility in economic policy around the globe. The rapid integration of global economies cautioned them against their raising stakes in global affairs. Information communication technology enabled lightening fast transmission of information and real time response by countries to any global event has also played a significant contributory role. Since the turn of the century, the global uncertainty has unprecedently increased owing to the occurrence of high-profile global events. For instance, events like immigration crisis, Britain’s delinking itself from European Union, political turmoil in Turkey (Turkish coup 2016), Brazil (impeachment of Dilma Rousseff in 2015) and South Korea (removal of president Park Geun-hye in 2017) and rise of populist political forces in some countries have shook all echelons of global world.

Uncertainty is a latent variable and carries an inherent difficulty of measurement. To this, burgeoning literature has produced range of alternate proxies or indicators based on either surveys or time series models. Survey based proxies primarily include ‘Forecast Disagreement’ (disagreement among the professional economic forecasters), finance-based VIX index and news publications based uncertainty index. Various studies have utilised time series modelling to obtain measures for uncertainty. For this, researchers frequently turn to stochastic volatility and generalized autoresgressive conditional heteroskedasticity (GARCH) models and have produced comparable results relative to the survey-based measures.

Taking guidance from earlier literature, we proxy our policy uncertainty by Global Economic Policy Uncertainty (GEPU) index developed by Davis and Steven (2016). GEPU index is GDP-weighted average of EPU indices of 20 individual countries which account for more than 70% of the global output. EPU itself is constructed on the basis of frequency with which a trio of terms pertaining to economy, policy and uncertainty appears in 10 leading newspapers of a country and is then normalised to a mean of 100.

Given the objectives of this study, using GEPU seems logical and reasonable for number of reasons as compared to the other finance-based measures having limited applicability. Finance-based measures are indirectly related to economy and the link between stock market volatility and economic growth is not clear (Shiller, 1981). In a similar vein, disagreement among forecasters captures forecast dispersion rather than uncertainty. As argued by Rich et al. (2012), differentiating between forecast dispersion and uncertainty makes forecaster disagreement a poor representation of uncertainty. Moreover, GEPU has certain advantages over the other measures of policy uncertainty. Apart from being extremely timely, this measure captures broad range of uncertainties including finance-based stock market volatility.

IV. Methodology

Specification of the Model

The present study attempts to explore the relationship between trade openness, global uncertainty and economic growth in India. To fulfil our objective, we followed the existing literature for model specification and identification of variables. We began our analysis with the following general formulation Equation (1) which is a modified version of Solow’s (1956) growth model. Equation (1) establishes relationship between the economic growth rate (EG) and the factors determining it:

E G = f (G C F, I N F, T R A D E, U N C E R)

(1)

Specifically, the equation can be written as:

E G_{t} = β_{0} + β_{1} G C F_{t} + β_{2} I N F_{t} + β_{3} T R A D E_{t} + β_{4} U N C E R_{t} + v_{t}

(2)

where EC, the dependent variable, is the economic growth which is proxied by quarterly GDP growth rate; GCF represents gross capital formation, a proxy for rate of investment; INF is the rate of inflation prevailing in the country thereby representing the domestic price stability; TRADE indicates the degree of openness measured by imports plus exports as a ratio of GDP and UNCER is the uncertainty prevailing in the global environment. We have taken data on quarterly basis for the period of 2003 Q1–2019 Q3. The data on all the variables was retrieved from Organization for Economic Cooperation and Development (OECD) database except for global uncertainty which was collected from www.policyuncertainty.com.

Autoregressive Distributive Lag Approach of Cointegration

In order to investigate the impact of underlying variables on economic growth, various econometric models can be resorted to. These include residual based Engle and Granger (1987) model, Maximum Likelihood-based Johansen (1988, 1991) and Johansen and Juselius (1990). However, these tests are severely sensitive to the stationarity property of the variables and sample size. To overcome these issues, we followed the autoregressive distributive lag (ARDL) approach developed by Pesaran and Shin (1999) and popularised by Pesaran et al. (2001). Contrary to the other approaches, ARDL approach performs better irrespective of order of integration of variables. Unlike the traditional cointegration techniques it does not require system of equations to determine long and short-run estimates. Rather, both long as well as short-run coefficients are estimated simultaneously with a single reduced form equation (Shrestha & Chowdhury, 2007). Moreover, ARDL produces unbiased long-run estimates (Odhiambo, 2008). However, it underperforms when the data series are integrated of order I(2). We specify the ARDL(p q) form of Model (1) as fallows:

\begin{array}{l} E G_{t} = β_{0} + \sum_{i = 1}^{p} β_{1} Δ G C F_{t - i} + \sum_{i = 1}^{p} β_{2} Δ I N F_{t - i} + \sum_{i = 1}^{p} β_{3} Δ T R A D E_{t - i} + \sum_{i = 1}^{p} β_{4} Δ U N C E R_{t - i} \\ + γ_{0} E G_{t - 1} + γ_{1} Δ G C F_{t - 1} + γ_{2} Δ I N F_{t - 1} + γ_{3} Δ T R A D E_{t - 1} + γ_{4} U N C E R_{t - 1} \\ + v_{t} \end{array}

(3)

Where all the variables are same as defined abovein equation (1). Δ is the first difference operator, p & q is the optimal lag length, β₁, β₂…. β₂ are the short-run responses while as γ₀,γ₁….γ₄ represent the long-run responses and υ_t is the error term. The error correction form of the Model (1) is expressed as follows:

\begin{array}{l} E G_{t} = α_{0} + \sum_{i = 1}^{p} α_{1} Δ G C F_{t - i} + \sum_{i = 1}^{p} α_{2} Δ I N F_{t - i} + \sum_{i = 1}^{p} α_{3} Δ T R A D E_{t - i} + \sum_{i = 1}^{p} α_{4} Δ U N C E R_{t - i} \\ + \partial_{0} E C M_{t - 1} + μ_{t} \end{array}

(4)

where ECM is the error correction term with one period lag, ∂₀ is the coefficient of the error correction term, and all other variables are same as defined as in Equation (2). Model 2 is similar to that of Model 1 with an addition of another variable. The added variable is an interaction of trade openness and global uncertainty included with an intent to capture the mediating role of trade openness in growth–uncertainty relationship. However, we did not produce the mathematical equation of Model 2 here.

Before application of ARDL model, the integration of all the variables must be checked. For that purpose, standard Augmented Dickey–Fuller (ADF) and Phillips–Perron (PP) test are employed. To determine the existence of long-run relationship or cointegration in Equation 3, bounds test is used. Bounds test determines the [non]existence of any cointegrating relationship among the variables using F-statistics. Null hypothesis is rejected [accepted] if F-statistic turns out to be greater [lower] than upper [lower] bounds’ critical values. Values of F-statistic lying between the two bounds are considered to be inconclusive. Cointegration testing is followed by selection of optimal lag length through various information criterion like Akaike Information Criteria (AIC), Bayesian Information Criteria (BIC), R² and Hannan–Quinn (HQ).

V. Results and Discussion

Fundamental to the time series analyses is the stationarity of the data. Stationarity of data has the property that the mean, variance and autocorrelation structure do not change over a period of time.

Precisely, it means a flat looking series with constant variance, without trend and no

periodic fluctuations (Enders, 2008). In this study, we applied ADF and PP unit root tests. The results pertaining to these tests are given in Table 1.

Table 1.

Unit Root Tests.

Variable	ADF at Levels		ADF at Difference		PP Test at Levels		PP Test at Difference
Variable	Static	Prob.	Static	Prob.	Static	Prob.	Static	Prob.
GCF	–5.190	0.0000	–	–	–5.590	0.0000	–	–
EG	–4.498	0.0000	–	–	–4.439	0.0000	–	–
INF	–0.555	0.4728	–3.626	0.0005	–0.548	.475	–8.001	0.0000
UNCER	0.107	0.7133	–10.837	0.0000	0.0791	0.704	–12.217	0.0000
TRADE	0.211	0.7447	–7.264	0.0000	0.1789	0.7350	–7.2526	0.0000

Source: The authors.

Notes: ADF: Augmented Dickey–Fuller; EG: Growth rate of GDP; GCF: Gross capital formation; INF: Inflation rate; PP: Phillips–Perron; TRADE: Trade openness; UNCER: Global uncertainty index

Table 1 indicates that gross capital formation (GCF) and growth are I(0) that is stationary at levels as the p-value stood under the threshold of 0.05. The other three variables (INF, uncertainty and trade) are non-stationary at levels as indicated by their p-values which stood higher than the threshold of 0.05. However, after differencing, the stationarity assumption is upheld in all the variables. All variables are I(0) at the first difference as indicated by their significantly lower p-value than its threshold of 0.05. Both ADF and PP tests verified the presence of unit-root in the dataset and present a mixed picture, thereby validating the application of ARDL approach.

Estimations by applying ARDL approach is divided into two stages. In the first stage, one must check if long-run equilibrium relationship between the variables exists or not. For that purpose, we applied the bounds F-test which hypothecates for non-existence of long-run relationship. Table 2 reports results of F-test on joint significance of long run coefficients.

Table 2.

ARDL Bounds Test: Null Hypothesis: No Long-run Relationships Exist.

Model 1	Test Statistic	Sig. (%)	I(0)	I(1)
F-statistic	5.828462	10	2.45	3.52
K	4	5	2.86	4.01
		2.5	3.25	4.49
		1	3.74	5.06
Model 2
F-statistic	5.062746	10	2.26	3.35
K	5	5	2.62	3.79
		2.5	2.96	4.18
		1	3.41	4.68

Source: The authors.

Note: ARDL: Autoregressive distributive lag

H₀: γ₀ = γ₁ = γ₂ … = γ₈ = 0 against the

H₁: H₀ is not true.

Table 2 reveals that the test statistic in both the models exceeds the I(1) critical value bound at all standard levels of significance. This provides a confirmation of presence of long-run relationship between the GDP growth and other independent variables. The confirmation of equilibrium relationship between the variables leads to the second stage. The second stage involves estimation of long run as well as the short-run responses of economic growth to independent variables especially trade openness and global uncertainty. We expect a positive relationship between GCF and economic growth and a negative relationship between the INF and GDP growth rate. However, we make no hypothesis regarding the effect of trade openness and global uncertainty on GDP growth rate. Table 3 presents the results of both the models through ARDL approach.

Table 3.

ARDL Estimates: Dependent Variable EG.

Long-run Coefficients	Variable	Model 1		Model 2
Long-run Coefficients	Variable	Coefficient	Prob.	Coefficient	Prob.
	GCF	0.010 (0.003)	0.0046	0.003 (0.0009)	0.0003
	INF	0.002 (0.001)	0.0425	0.004 (0.001)	0.0008
	UNCER	–0.003 (0.008)	0.6747	–0.049 (0.064)	0.4466
	TRADE	0.074 (0.082)	0.3698	0.499 (0.680)	0.4668
	INTER	–	–	0.057 (0.142)	0.6894
	C	0.035 (0.062)	0.5755	0.331 (0.311)	0.2928
Short-run coefficients
	D(EG(–1))	0.557 (0.177)	0.0030	0.421 (0.171)	0.0183
	D(GCF)	0.005 (0.001)	0.0003	0.005 (0.0009)	0.0000
	D(INF)	–0.004 (0.003)	0.1874	–0.003 (0.003)	0.3702
	D(INF)	0.002 (0.004)	0.4891	0.003 (0.004)	0.4116
	D(INF)	–0.007 (0.003)	0.0299	–0.008 (0.003)	0.0173
	D(UNCER(–1))	–0.004 (0.011)	0.0752	–0.045 (0.024)	0.0729
	D(TRADE)	0.551 (0.248)	0.0319	0.677 (0.906)	0.0589
	D(INTER)	–	–	0.114 (0.039)	0.0065
	(ECM(–1))	–1.439 (0.248)	0.0000	–1.354 (0.213)	0.0000
R² Adjusted R² F-statistic Prob (F-statistic)		0.865020 0.806960 3.362006 0.000617		0.774783 0.606785 3.229157 0.000826

Source: The authors..

Notes: Standard errors are given in parenthesis.

ARDL: Autoregressive distributive lag; EG: Growth rate of GDP; GCF: Gross capital formation; INF: Inflation rate; TRADE: Trade openness; UNCER: Global uncertainty index.

Table 3 reveals that most of the variables are statistically significant and have expected signs. The upper panel shows the long-run results and lower panel shows the short-run results. More precisely, GCF has as expected positive and statistically significant coefficient. GCF contributes positively to enhance the economic performance in India. The point estimates of GCF is 0.010 and 0.003 with p-values 0.0046 and 0.0003, respectively. The economic interpretation of the results is that with every 10 incremental units of capital formation, economy grows by 0.3 to 1 units. The coefficient of GCF is qualitatively similar in Model 2, which shows similar effects on economic growth. However, the inclusion of interaction term has quantitatively affected the coefficient of GCF. The magnitude of its impact decreases from 0.10 to 0.003 units per incremental unit change in capital formation. Though impact is shown marginally low yet it is statistically significant. In the short run also, as indicated by the second row of the lower panel, GCF contributes positively to the economic growth in India. This could suggest that the investment policies adopted by the government in India have been beneficial to its economic progress.

Contrary to the expectations, INF have significant and positive but negligible impact on EG in the long run as shown in Table 3. INF is generally believed to be retarding the growth. However, a mild INF acts as an encouragement to the producers to produce more. Since the variability in consumer prices in India have stabilised over time with only occasional upward spikes, INF is unlikely to have recessionary effect on growth. These results suggest that INF in India has not reached a threshold level beyond which INF affects growth negatively. We note that our results are in line with the study (Khan, 2001) which estimated threshold level of INF for a panel of 140 countries over the period of 1960–1998. Khan (2001) argued that INF rate beyond 11–12% only exerts negative impact on growth in developing countries. Ghosh and Phillips (1998) while investigating non-linear INF–growth nexus also reported similar results.

As indicated in the upper panel of Table 3, uncertainty shares an expected negative correlation with economic performance of India. The estimated coefficient of uncertainty (UNCER = −0.003) is both economically negligible and statistically insignificant meaning thereby that the shocks if any are temporary and does not have any severe long-term impact on Indian economy. These results are plausible because over the years, fluctuations in Indian economy have not been linked closely with the global uncertainty. This is evident from the remarkable growth rate of Indian economy despite the buildup in global uncertainty following financial crisis in 2008. Growth rate was higher than the projections of IMF and World Bank at 6.5% in 2008 and reached 8.6% and 9.3% in the following two years. This indicates how resilient Indian economy is against the external shocks. The immunity of Indian economy to the global shocks is primarily because of its two important and fundamental characteristics: (1) the growth in India is almost entirely driven by the domestic consumption demand with only a little contributory role played by other factors and (2) the lower dependence on external sector. Though Indian economy has progressively moved towards neutral trade regime by eschewing tariffs and other barriers yet it has to fully integrate with the global economy. Contrary to this, uncertainty affects growth rate both negatively and significantly in the short run. As shown by the seventh row of the lower panel, the coefficient pertaining to uncertainty is significant in both the models. This means that the uncertainty drives the business cycles in India.

With regard to the trade openness, the results indicate that trade openness does not determine growth in the long run. The estimated coefficients in both the models are positive but insignificant. This finding lends support to Ulasan (2015), who used four alternative proxies of trade openness and concluded that there exists no robust trade–growth relationship. However, Greenaway and Wright (2002) argued that in the long-run trade does not determine the economic growth but only plays a transitionary role. Using a core growth model with an addition of alternate measures of liberalisation, they stated that there exists a medium-run relationship between trade and economic growth. The short-run coefficients of trade openness given in the lower panel of Table 3 are both positive as well as statistically significant at standard levels of confidence. These results partially support the conclusion of Greenaway and Wright (2002).

Central to this study is the joint effect of uncertainty and trade openness captured through their interaction. In Model 1, we deliberately excluded their joint effect in order to examine the individual effect of trade openness and uncertainty. Upon its inclusion, the long-run coefficient of trade openness improves but still remains insignificant. The coefficient of interaction term is positive but lacks significance. The economic interpretation is that the joint increment in trade openness and uncertainty by 10 units improves the overall economy by four units. Moreover, in the short run, coefficient of the interaction term is positive as well as statistically significant. Since uncertainty has a significant negative impact on short-run economic growth, positive and significant coefficient of interaction term depicts the ability of trade openness to contain the negative consequences of global uncertainty. The major takeaway is that trade openness mitigates the negative consequences of global uncertainty both in the long- as well as short run. To the proponents of protectionism, which suggest closed economic structure as a shield against global uncertainty, this finding may seem counterintuitive. This finding compare well with the empirical evidences reported by Cavallo and Frankel (2008), who argued that trade acts as an absorbent to the external shocks.

The confirmation of cointegration among variables also comes from the coefficient of error correction term (ECM(−1)). The coefficients are both statistically significant at 1% level of significance with the expected negative sign. This indicates that the correction of any short-term fluctuation in response to a shock takes slightly more than two months (62 and 66 days¹) at a speed of 145% and 135%, respectively.

While these results may seem to be influenced by independence, variance and covariance properties of the regression error term, it is customary to perform certain residual diagnostic testing. Accordingly, we performed Breusch Pagan/Cook–Weisberg test for heteroskedasticity and Breusch–Godfrey serial correlation Lagrange Multiplier (LM) test. These results are presented in Table 4.

Table 4.

Residual Test.

Test	Model 1		Model 2
Test	Test Statistic	P-value	Test Statistic	P-value
Heteroskedasticity test: Breusch Pagan/Cook–Weisberg	0.913625	.5114	0.594419	.8285
Breusch–Godfrey serial correlation LM test	0.857851	.2962	0.779398	.3213
Durbin–Watson stat	2.055521		1.964671

Source: The authors.

The results displayed by the two models in Table 3 pass all diagnostic tests. Table 4 reports the residual diagnostic tests of serial correlation and heteroscedasticity. Under the null hypothesis of serially uncorrelated residuals, the p-value associated with F-statistic of serial correlation LM test is found to be 0.2962 and 0.3213 for Models 1 and 2, respectively. Therefore, we fail to reject this null hypothesis and conclude that residuals are serially uncorrelated. In a similar vein, under the null hypothesis of homoscedasticity of residuals, the p-value associated with F- statistic of Breusch Pagan/Cook–Weisberg test for heteroskedasticity is found to be 0.5114 and 0.8285for Models 1 and 2, respectively. Again we also fail to reject this null hypothesis and conclude that residuals are homoscedastic. Thus, we can conclude that the results of our study are robust and consistent. Further, to test the stability of the models we performed CUSUM and CUSUM Square test and the graph is plotted in Figures 1 and 2.

Again both the models passed the stability test as given by CUSUM and CUSUMQ. The graphical representation of stability tests are shown in Figures 1 and 2. The plot of cumulative sum of recursive residuals and squared residuals indicates that the parameters are stable over the sample period. Overall, Table 4 and Figures 1 and 2 indicate that the model has desirable statistical and theoretical properties and can be used for policy analysis.

VI. Concluding Remarks

In summary, the present study explored the long- and short-run effects of trade openness and global uncertainty on Indian economic performance from 2003Q1 to 2019Q3 using recently developed econometric technique called ARDL bounds testing approach. Specifically, the study focuses on two issues: individual examination of trade openness and global uncertainty for their impact on economic growth and if trade openness mitigates or exacerbates the effect of external shocks on Indian economy. Overall the results suggest that the evidences of existent long-run trade–growth and uncertainty–growth link are weak. Controlling for other plausible determinants of growth, we find only capital formation as significant predictor of growth in the long run.

The effect of trade openness and global uncertainty are significantly felt in the short run. Their effect appears not only qualitatively significant but also economic viable. Our striking finding is that trade openness does not increase India’s vulnerability to external shocks arising from global events. Rather trade openness performs the role of an absorbent to mitigate the negative consequences of global uncertainties.

While due care has been taken to ensure the empirical results are robust and reliable, yet this study suffers from certain limitations. The model may be underspecified due to the non-availability of high frequency data on various factors affecting growth. Once data become available on factors like capital flows and other significant determinants of growth, it would be interesting to test if the results change after taking such variables into consideration.

Based on the results, those at the helm of the affairs should draw a comprehensive strategy to enhance the contribution of investment to long-term economic development. International trade shows the signs of potential growth effect. Policy makers should follow reformist ideology to redraw the trade policy focusing upon increasing the share of capital and heavy industrial goods in the export basket.

Footnotes

Availability of Supporting Data

The data is readily available with the authors and can be provided upon request.

Declaration of Conflicting Interests

Ethical Standards

This article contains no malicious content.

Funding

The authors received no financial support for the research, authorship and/or publication of this article.

ORCID iDs

Mehraj Ahmad Sheikh

Mushtaq Ahmad Malik

Note

References

Alessandri

, & Mumtaz

(2014). Financial regimes and uncertainty shocks (School of Economics and Finance Working Paper No. 729). Queen Mary University of London.

Baker

S. R.

, Bloom

, & Davis

S. J.

(2013). Measuring economic policy uncertainty (Chicago Booth Research Paper 13–02, Vol. 22, p. 81). Stanford University.

Balassa

(1986). Policy responses to exogenous shocks in developing countries. The American Economic Review , 5, 75–78.

Bhagat

, Ghosh

, & Rangan

S. P.

(2013). Economic policy uncertainty and economic growth in India. SSRN Electronic Journal , 51(35). 10.2139/ssrn.2246882

Bijsterbosch

, & Guérin

(2013). Characterizing very high uncertainty episodes. Economics Let-ters , 121(2), 239–243.

Bloom

(2009). The impact of uncertainty shocks. Econometrica , 77(3), 623–685.

Caggiano

, Castelnuovo

, & Groshenny

(2013). Uncertainty shocks and unemployment dynamics: An analysis of post-WWII US recessions . University of Padova.

Carroll

C. D.

(1997). Buffer-stock saving and the life cycle/permanent income hypothesis. The Quarterly journal of economics , 112(1), 1–55.

Cavallo

E. A.

, & Frankel

J. A.

(2008). Does openness to trade make countries more vulnerable to sudden stops, or less? Using gravity to establish causality. Journal of International Money and Finance , 27(8), 1430–1452.

10.

Clark

P. B.

(1973). Uncertainty, exchange risk, and the level of international trade. Western Economic Journal , 11(3), 302–313.

11.

Coe

D. T.

, & Helpman

(1995). International R&D spillovers (Working Paper No. 4444). http://www.nber.org/papers/w4444

12.

Colombo

(2013). Economic policy uncertainty in the US: Does it matter for the Euro area? Economics Let-ters , 121(1), 39–42.

13.

Condon

, Corbo

, & De Melo

(1985). Productivity growth, external shocks, and capital inflows in Chile: A general equilibrium analysis. Journal of Policy Modeling , 7(3), 379–405.

14.

Davis

S. J.

(2016). An Index of Global Economic Policy Uncertainty . Macroeconomic Review, October.

15.

Dollar

(1992). Outward-oriented developing economies really do grow more rapidly: Evidence from 95 LDCs, 1976–1985. Economic Development and Cultural Change , 40(3), 523–544.

16.

Enders

(2008). Applied econometric time series . John Wiley & Sons.

17.

Frankel

J. A.

, & Romer

D. H.

(1999). Does trade cause growth? American Economic Re-view , 89(3), 379–399.

18.

Froning

D. H.

(2000). The benefits of free trade: A guide for policymakers. Heritage Foundation Backgrounder , No. 1391.

19.

Fujita

(2007). Toward a new modeling of international economics: An attempt to reformulate an international trade model based on real option theory. Physica A: Statistical Mechanics and its Applications , 383(2), 507–512.

20.

Ghosh

, & Phillips

(1998). Warning: Inflation may be harmful to your growth. Staff Pa-pers , 45(4), 672–710.

21.

Greenaway

, Morgan

, & Wright

(2002) Trade liberalisation and growth in developing countries. Journal of Development Economics , 67(1), 229–244.

22.

Greig

, Rice

, Vehbi

, & Wong

(2018). Measuring uncertainty and its impact on a small open economy. Australian Economic Review , 51(1), 87–98.

23.

Hooper

, & Kohlhagen

S. W.

(1978). The effect of exchange rate uncertainty on the prices and volume of international trade. Journal of International Economics , 8(4), 483–511.

24.

IMF. (2013). Spillover report—Analytical underpinnings and other background . IMF Spillover Report. International Monetary Fund.

25.

Johansen

(1988). Statistical analysis of cointegration vectors. Journal of Economic Dynamics and Control , 12(2–3), 231–254.

26.

Johansen

(1991). Estimation and hypothesis testing of cointegration vectors in Gaussian vector autoregressive models. Econometrica: Journal of the Econometric Society , 59(6), 1551–1580.

27.

Johansen

, & Juselius

(1990). Maximum likelihood estimation and inference on cointegration—With applications to the demand for money. Oxford Bulletin of Economics and Statistics , 52(2), 169–210.

28.

Jurado

, Ludvigson

S. C.

, & Ng

(2013). Measuring uncertainty . Mimeo, New York University.

29.

Khan

M. S.

, & Sinhadji

A. S.

(2001). Threshold effects in the relationship between inflation and growth. IMF Staff Papers , 48(1), 1–21.

30.

Kimball

M. S.

(1990). Precautionary saving in the small and in the large. Econometrica , 58(1), 53–73.

31.

Klobnera

, & Sakke

(2014). International spillovers of policy uncertainty. Economics Letters , 124(3), 508–512.

32.

Kraay

(1999). Exports and economic performance: Evidence from a panel of Chinese enterprises. Revue d’ Economique du Developpement , 2, 183–207.

33.

Krishna

, & Mitra

(1998). Trade liberalization, market discipline and productivity growth: New evidence from India. Journal of Development Economics , 56(2), 447–462.

34.

Leduc

, & Liu

(2012). Uncertainty shocks are aggregate demand shocks (Federal Reserve Bank of San Francisco Working Paper). https://www.frbsf.org/economic-research/files/wp12-10bk.pdf

35.

Lucas

R. E.

(1988). On the mechanics of economic development. Journal of Monetary Economics , 22(1), 3–42.

36.

Martin

, & Rey

(2006). Globalization and emerging markets: With or without crash? American Economic Review , 96(5), 1631–1651.

37.

Moore

(2016). Measuring economic uncertainty and its effects (Research Discussion Paper 2016-01). Economic Research Department Reserve Bank of Australia.

38.

Odhiambo

N. M.

(2008). Financial depth, savings and economic growth in Kenya: A dynamic causal linkage. Economic Modelling , 25(4), 704–713.

39.

Pesaran

M. H.

, Shin

, & Smith

R. J.

(2001). Bounds testing approaches to the analysis of level relationships. Journal of Applied Econometrics , 16(3), 289–326.

40.

Rich

, Song

, & Tracy

(2012). The measurement and behavior of uncertainty: Evidence from the ECB survey of professional forecasters [Federal Reserve Bank of New York Staff Reports No 588]. Federal Reserve Bank of New York.

41.

Romer

P. M.

(1986). Increasing returns and long-run growth. Journal of Political Economy , 94(5), 1002–1037.

42.

Rose

(2005). A reason why countries pay their debts: Renegotiation and international trade. Journal of Development Economics , 77(1), 189–206.

43.

Sachs

J. D.

(1985). External debt and macroeconomic performance in Latin America and East Asia. Brookings Papers on Economic Activity , 2, 523–564.

44.

Shiller

R. J.

(1981). Do stock prices move too much to be justified by subsequent changes in dividends. The American Economic Review , 71(3), 421–436.

45.

Shrestha

M. B.

, & Chowdhury

(2007). Testing financial liberalisation hypothesis with ARDL modelling approach. Applied Financial Economics , 17(18), 1529–1540.

46.

Solow

R. M.

(1956). A contribution of the theory of economic growth. Quarterly Journal of Economics , 70(1), 65–94.

47.

Stokey

N. L.

(1988). Learning by doing and the introduction of new goods. Journal of Political Economy , 96(44), 701–717.

48.

Ula͑an

(2015). Trade openness and economic growth: Panel evidence. Applied Economics Letters , 22(2), 163–167.

49.

Were

(2015). Differential effects of trade on economic growth and investment: A cross country empirical investigation. Journal of African Trade , 2(1–2), 71–85.