Impact of Public and Private Infrastructure Investment on Economic Growth: Evidence from India

Abstract

Literature presents contradictory views regarding the impact of public and private investment on the economic growth of a country. India being a developing country, where the major share of investment is by public sector, the question which props up is what among public and private investment is contributing more towards the economic growth of the country. In this framework, the gross domestic product (GDP) can be fairly explained as a function of public infrastructure investment and private infrastructure investment. Johansen’s co-integration was used to test the long-run relationship between the variables over the period from 1961–1962 to 2016–2017. A vector error correction model (VECM) along with an impulse response function and variance decomposition analysis was done to measure the impact of public infrastructure investment and private infrastructure investment on the GDP. Based on the empirical evidence discussed earlier, it was evident that both public and private infrastructure investments have a significant impact on the economic growth of the nation. Findings which came up in this study correlate to majority findings of past literature that, when compared with public investment, it is private investment which is capable of giving a better impetus to economic growth.

Keywords

Public infrastructure investment private infrastructure investment GDP Indian economy

JEL Classification: H54, R42

1. Background

In research, the relative impact of public and private investments on economic growth always gained attention of scholars. As there was a paramount shift of emphasis from public sector to private sector in the global scenario, evidence from India was visible since the beginning of the New Economic Policy–1991. The private sector leadership and the curtailment of the relative importance of public sector have invited appraisals as well as criticisms from all over the country, although its long-term impact is still questioned. Even now, literature presents contrary views when it comes to the impact of public and private investments on economic growth. It is well documented that public and private investments have a significant positive impact on the economic growth of any country. There is absolute concurrence to the fact that whichever be the source of spending—public or private investment—it aids in lifting up the GDP of an economy (Aziz, 2014). But the relative share of public and private investments matters the most.

While going through the related literature, a bag full of conflicting results on the importance of public and private investments can be found. Some studies emphasised more on the relative role of public sector investments, while others stressed on private sector investments. This may be due to the difference in the economy or the data period under study, or it may even be due to the difference in methodology that was adopted in the study. Right from the time of the Great Depression of the 1930s, Keynes advocated the increased role of government expenditure in stimulating the economic growth rate. In the work of Munnell (1990a), it has been found that public capital has a greater impetus to economic growth. It was also argued in the study that private investment, which is made from a limited pool of savings by private individuals, can be carried out only by replacing other capital projects. The limitation in the quantity of the funds is the main reason that outshines public investment over the other. Aschauer (1989) also in his work glorified public investment as more beneficial to economic growth. His work is considered as one of the most remarkable ones in this area.

But to our surprise, there are studies wherein public capital has been proved to have a negative impact on output. The work of Evans and Karras (1994) is one among them. Other notable study, which indicated the negative impact that public investment made on the relationship between private investment and economic growth, is by Epaphra and Massawe (2016). They state that the use of financial and physical resources by the public sector, which may otherwise be utilised by the private sector leads to crowd out of private investment in relation to public investment. Their study was conducted in Tanzania, which is a developing country. There are evidences that revealed the weak impact of public investment on economic growth in developing countries, while the same is reversed for developed countries. This idea is supported by Devarajan et al. (1996). However, there are evidences showing both public and private investments contributing equally to economic growth. The study performed by Afonso and St. Aubyn (2008) is one among them.

Why is India lying far behind when compared to the Western counties? Why is the growth of standard of living in India still lagging behind? The only answer to these questions is the lack of physical infrastructure available in the country. Enormous past infrastructure investments in those developed countries resulted in the improvement of the overall quality of living and thereby accelerated economic growth. It is widely acknowledged that infrastructure investment has a significant positive effect on economic growth from an early period itself. Economies, which have clocked faster economic growth, were always provided with adequate investment in infrastructure. Even after the drastic destructions that occurred during the Second World War, economies like China and Japan concentrated in rebuilding their infrastructure so that they could cope up and spur the development of their economy. Undoubtedly, in every economy, infrastructure acts as a catalyst to the overall development.

In the Global Competitiveness Report of the World Economic Forum (2017–2018), India has been ranked 66th out of the 137 countries in overall infrastructure and could score only 4.2 points out of 7. Being the seventh largest economy in the world, these figures are really disappointing. The lack of infrastructure is considered as the main cause for India’s diminishing GDP by 1–2 per cent every year (Nataraj, 2014). Over the past few years, investment in infrastructure has increased considerably by the Government of India (Neeraj, 2013). The government’s focus has shifted to building infrastructure from the Tenth Five Year Plan itself. A total amount of ₹5,970 billion has been allocated for the development of infrastructure in the financial budget of 2017–2018. But the government alone cannot bridge the huge infrastructure gap that has existed in India due to the ever-increasing population. A parallel source of investment by the private sector must be there along with the government, in order to meet the rising needs of funds in infrastructure. The resource crunch arising from the need of additional investments in infrastructure can be solved to a great extent by attracting private finance to infrastructure. During the Twelfth Five Year Plan, half of the targeted investment in infrastructure was expected to be financed through private sector investment.

For catalysing private infrastructure investment in the Indian economy, the steps taken by the Reserve Bank of India (RBI) and Securities and Exchange Board of India (SEBI) are really appreciable. Long-term loans for infrastructure development provided by banks have been liberalised by the RBI. Moreover, banks have been motivated to invest in Special Purpose Vehicle (SPV) in the private sector, specially designed for carrying out infrastructure projects. SEBI has also provided many relaxations in terms of minimum subscription, price fixing, etc., especially for infrastructure companies. An Infrastructure Development Finance Company (IDFC) was also set up in 1997 for providing funds, advisory services and asset management services for infrastructure projects in India. Despite these activities, the private investment inflow into the infrastructure sector was found to be inadequate (Varshney, 2008). India being a developing country, where a major share of investment is made by the public sector, the question which props up is what among public and private investments is contributing more towards the economic growth of the country. This puzzle has not been solved in the infrastructure sector too. There is still no agreement regarding the superiority of public or private investment over others in the sector. Being a crucial area and involving a substantial amount of investment, which sector is investing in infrastructure is a pivotal question.

Usually, in developing countries, it is considered as an obligation of the public sector to investment in infrastructure. But at the same time, the relative importance of the private sector in today’s growing economy cannot be neglected. We can also see that a substantial weighting is still provided to boost private investment in the economy. The world’s economic infrastructure has witnessed a transfer of ownership over the past decade. A major share of world infrastructure is in the hands of private investors. In India, immediately after the independence, it was the obligation of the government to invest in the necessary sectors of the economy. But now the scenario has undergone a drastic change. The traditional outlook that it is the sole duty of the government to make investments for the well-being of society has gradually disappeared. Now, it has become necessary to combine private investment with public investment to meet the large investment gap, and disclosing the fact that the government cannot handle the whole responsibility alone.

India is a country where the government and private parties join hands in investing in vital areas of the economy. Infrastructure sector is no longer an exception. Both the public and private sectors contribute to the growth of infrastructure in the country. In this scenario, which, among public or private investment, is contributing more towards the economic growth in India is a worthwhile question. This question gains importance when the relationship between public investment and private investment does not always become complementary. The impact that both public infrastructure investment and private infrastructure investment create on the economic growth will also be different in this case. However, it is surprising to note that no study has been conducted till date which provides a satisfactory answer regarding this. Against this backdrop, this study attempts to find out the relative contribution of both public and private infrastructure investments to the economic growth of India. This study is a novel attempt to explore the superiority and impact that public and private investments have on the infrastructure sector and to the economic growth of a developing country like India.

2. Review of Existing Literature

A good number of studies have been conducted regarding the relative contribution of public or private investment on economic growth of developing and developed countries. Many of them came up with conflicting findings. These studies can be bifurcated into those which highlighted the effect of public investment to the economic growth vis-à-vis private investment to the economic growth. As studies pertaining to infrastructure are rarely found, general studies on impact of public and private investments on economic growth are discussed. It is important to note that, while probing through the existing literature regarding the relative impact of public investment and private investment over the economic growth of countries, the results may vary subject to regional and temporal variations.

2.1 Empirical Studies in Developed Countries

Out of many studies conducted in the USA, the study of Evans and Karras (1994) and Tatom (1991) stood out, as they both pointed out the negative effect of public capital in the productivity of US economy. Evans and Karras (1994) investigated the impact of government capital and other government activities on the private production in the US economy from 1970 to 1986. Data were analysed using Cobb–Douglas along with translog aggregate production function. The empirical findings showed that apart from education services provided by the government, no other government activities were found to be productive. Moreover, the productivity of government capital was also found to be negative. The study also suggested that the government investment must be underprovided when cost of such activities become more than their contribution to private output. Unless it becomes empirically proved that government investment creates direct non-market consumption expenditure, pricing and carrying out of existing government activities must be prioritised than providing more investment. Tatom (1991) also investigated about the role of public capital in the private sector performance of the US economy, using Cobb–Douglas production function and co-integration method, by using annual data from 1949 to 1989. Unlike Evans and Karras (1994), to overcome the problem of spurious results, the study considered stationarity of data and the effect of energy price levels. Based on the findings of the study, it was proven that increased government capital spending no longer impacts private sector output, productivity and private capital formation. The elasticity of private output to public capital has fallen from 30–40 per cent to 13 per cent when stationarity of data is considered .

While the findings of these studies pointed out to the inefficiency of public capital investment in the US economy, Zou (2006) attempted to compare the interaction between public investment, private investment and GDP growth of the USA with Japan. As there were differences in some features of data, annual data from 1958 to 1997 were analysed using Generalised Method of Moments for Japan and ordinary least squares (OLS) for the USA. From the analysis, it was found that there exists a long-term relation between public investment, private investment and GDP of both countries. But the relative impact of public and private investments was different due to variances in economic, political and even historical factors. In Japan, both public and private investments contributed equally to the growth of GDP, whereas, in the USA, it is the private investment that played a major role than public investment. This was due to the fact that being a capitalist country, private sector seemed to be an overwhelming force in the USA. But for the Japanese market system, the role of the government was equally important.

But there were studies that proved that private investment was no longer a strong force than public capital even in capitalist countries. An important fact among them is the prominent study conducted of Aschauer (1989). He studied the impact of public sector capital accumulation and flow of government expenditures on productivity growth in the US economy. Annual data from 1949–1985 was analysed using the regression method. The findings of the study states that deemphasising the role of public finance in the economy has adversely affected the productivity of the economy. Public investment decisions, especially in non-military sectors such as highway, water systems, etc., have a substantial weight in the economy. Although public sector deficits have an impact on real interest rates and private investment decisions, they have to be given comparable or even dominating significance for improved productivity in the economy. This study of Aschauer has claimed widespread attention and paved the way for many such studies. But it was not free from criticisms too. Sturm and de Haan (1995) claimed that empirical findings of studies including Aschauer (1989) that public capital is linked to productivity is not well founded due to the non-stationarity of data. For this, in his study, the impact of public capital stock on productivity in the USA and Netherlands from the 1949–1985.was analysed using the OLS method along with Engle–Granger test. After first differencing, the result was not justifiable according to economic theories as private capital elasticity and labor elasticity was negative in both USA and Netherlands.

There were also some other studies with identical findings in the US economy that praised the role of public investment in the productivity of US economy. Munnell (1990) explored whether the growth of public capital along with growth of private capital and labour made an impact from 1949 to 1987. Empirical findings suggest that a shortfall in public investment will lead to lagged growth of labour productivity. More specifically, decrease in the public infrastructure was identified as a crucial factor for this. Increased public spending will not only eliminate this drag but also will increase the pubic capital–labour ratio and eventually labour productivity. Although increase in public spending will create temporary pressure on the government, failing to do so will result in serious decline in the country’s productivity. Pereira and de Frutos (1999) also studied the impact of public capital on private sector variables in the US economy. Data during the period from 1956 to 1989 were analysed using co-integration method, Vector Auto Regression (VAR) and impulse response function. Empirical results suggest that pubic capital positively influenced lagged output but negatively influenced lagged employment. Moreover, there was a crowd-in relationship existing between public capital and private capital. The study also recommended not using public capital as a counter-cyclical tool. This is because public investment actually occurs at the peak of a business cycle, when investment is no longer needed. Due to this, it rarely meets the needs of long-term needs of infrastructure investment.

Ram and Ramsey (1989), while estimating the aggregate production function for private output in the USA, investigated the effect of government capital on the same. Data from the period 1948–1985 were studied using Cobb–Douglas production function. The findings of the study proved that there was a positive impact concerning the government capital when it was used for improving the private output of the US economy. When the government capital was segregated into federal capital and state and local capital, the latter appeared to be more influential and statistically significant. The study not only reaffirmed the role of the government in increasing output but also found that state and local capital were an important source behind it. Lynde and Richmond (1992) also estimated the impact of public capital on the productivity of US non-financial corporate sector. Annual data ranging from 1958 to 1989 were analysed using translog cost function and OLS method. The findings of the study show that public capital is an essential input in the production function, and the marginal productivity is also positive. In addition to this, the findings also established a complementary relationship between public and private capital instead of being substitutes. The public capital in infrastructure also plays a substantial role in private sector productivity. Although in most countries there is a natural monopoly of public sector in infrastructure, private sector participation is also necessary. Lynde and Richmond (1993) later analysed the impact of public capital in the productivity and output levels of the US economy. The findings of the study show that a decrease in the public capital–labour ratio led to the decline in productivity of the economy. While majority of previous literature neglected the role of public capital in the USA, this study stood hand-in-hand with other rare works like Aschauer (1989) and Munnell (1990).

Erenburg (1993) examined the relationship between public and private investments in the US economy. Annual observations from 1947 to 1985 were estimated using simple rational expectations model. The findings of the study exhibit that there existed a statistically significant positive relationship between public capital investment and private capital investment in the US economy. This showed a strong evidence for public capital must be explicitly taken into account while examining aggregate policy effects. Moreover, decrease in the spending of public capital, especially in infrastructure, leads to a decline in economic growth of the country. There are also evidences that showed that such decline in public capital also leads to a decline in the marginal productivity of the private sector investments. Crowder and Himarios (1997) analysed the relationship between public capital, private capital and real output level of post-war US economy. Data from 1947 to 1989 were analysed using Johansen’s co-integration method along with vector error correction model (VECM). The study attempts to validate whether the balanced growth restrictions, which explain the same stochastic trend of public capital, private capital and real output level, were true or not. The findings of the study revealed that the balanced growth restrictions by the neoclassical growth cannot be rejected. It also highlighted the relative importance of public capital’s productivity at the margin over private capital in contributing to economic growth. From any reasonable aggregate production function specification, public capital stock cannot be excluded.

Short-run and long-run relationships between public capital and output growth were investigated by Lighthart (2000) in Portugal during the period from 1965 to 1995. Data were analysed using Cobb–Douglas production function, Johansen’s co-integration and VAR. The findings suggested that public capital is an important variable in determining the output growth of Portugal. The findings are similar to other studies that suggest that the EU-supported public investment always has a positive effect on economic growth. While disaggregating data, public investment in sectors such as road, railway and airports shows more productivity in comparison to public investment in other sectors.

There were few studies that proved the equal importance of public and private investments. The study conducted by Afonso and St. Aubyn (2008) is one among them. They estimated the macroeconomic effects of public and private investments on the economic growth of 14 European countries along with Japan, Canada and the USA. By using annual data ranging from 1960 to 2005, VAR method with impulse response function was used for analysing the impact. The results showed the presence of positive effect of both public and private investments on the output of the countries. While the crowd-in effect of private investment on public investment can be generalised among the countries, the reverse crowd-in effect varies among the sample countries.

2.2 Empirical Studies in Developing Countries

To study the interrelationship among public investment, private investment and output of selected highly indebted poor countries, Belloc and Vertova (2004) used the Engle and Granger co-integration along with VECM and impulse response function. The findings depict that there existed a complementary relationship between public and private investment in six out of seven cases. A positive effect of public investment over the output was also strongly established from the findings of the study. In the case of Mexico, Nazmi and Ramirez (1997) analysed the impact of public and private investments on the economic growth of Mexico during the period between 1950 1990, using modified neoclassical production function, and the findings also exhibited the significant role of overall public investment in Mexico on its economic growth. It was also found that public investment does this at the private investment’s expense. This proves the crowd-out effect of public investment over private investment in Mexico. It was also interesting to note that in the study, both public and private investments had identical effect on economic growth. Evidence for the complementary relationship of public investment with private investment in developing countries was also an important finding in the study conducted by Serven and Solimano (1993). They investigated the contributing factors determining investment performance in 15 developing countries. Cross-country data from 1975 to 1987 were analysed using regression models. The findings also suggested that insufficient public investment, especially in infrastructure, along with private investment, can hamper the economic growth rate of the countries. But debt-financed public investment increases the cost of capital and thereby reduces the rate of return after deducting the tax of private investment. This has led to the decrease in private investment, eventually affecting the economic growth rate.

There is evidence proving the fact that the relationship between public and private investment need not always be complementary. In a study conducted by Coutinho and Gallo (1991), among 33 countries from 1970 to 1988, they evaluated the effect of public policies on private investment and to identify whether public investment increased the economic output or not. The results suggested that public and private investments act as a substitute and not complementary to each other in aggregate level. But public investment in infrastructure becomes a core ingredient that accelerates private investment in these countries. The importance of public investment in the infrastructure sector was also highlighted in many studies. Most importantly, the study conducted by Khan and Kumar (1997) examined the relative effects of public and private investments on the GDP per capita of 95 developing countries. Cross-country and panel data were collected and analysed using extended basic neoclassical model of growth. They found that public investment in infrastructure has a crowd-in effect for private investment in the long run. They also added that although both public and private investments influence the economic growth of these countries, their effect is subject to regional and temporal variations. Ramirez (1996) also proved in his study that public investment in infrastructure increases the marginal productivity and thus complements the private investment. In his study, the complementarily relationship of public and private investments and their contribution to economic growth in Mexico and Chile were analysed from 1940 to 1992 using linear growth model. The findings also suggest that both public and private investments have a positive effect on the economic growth of Chile and Mexico. This result warns the trend of reducing consumption expenditure of the government in the Mexican economy due to IMF norms as a danger that retards the growth of the economy.

In the Côte d’Ivoire economy, Bédia (2007) analysed the effect of public and private investments using auto regressive distributed lag (ARDL) model along with VECM. Data series over the period from 1961 to 2001 were studied and the findings suggested that in the short run, private investment has a great impact on economic growth in comparison to public investment. But the trend is just revered in the long run. In the long run, the public investment has much more impact than that of private investment in the Côte d’Ivoire economy. One of the suggestions also put forward by the study was to reconsider the decision of reducing investment by the government through privatisation and liberalisation processes. But this study conflicts with the findings of Ghali (1998) that made use of multivariate co-integration and VECM to evaluate the long-run effect of public investment on private investment and economic growth of Tunisia from 1963 to 1993. In the short run, public investment has a negative impact on private investment and no impact on economic growth. But in the long run, both private investment and economic growth are negatively affected by public investment. The study also suggested the rationalisation of public investment in Tunisia, in order to accelerate economic development.

The positive effect of private investment on the economic development was highlighted as the major finding in a number of studies. Ghura (1997) studied the determinants of economic growth of Cameroon and the influence of private investment on its growth. The study used Granger causality test along with OLS regression to analyse annual data ranging from 1963 to 1996. The findings depict that even though public and private investments have a positive effect on the economic growth, the influence of the latter was found to be more significant. This is because increase in government expenditure will lead to larger budget deficits, thereby adversely affecting economic growth. It is important to note that while reducing unproductive investments by the government, investment in infrastructure and human capital development must be safeguarded. In another study, Haque (2013) examined the relative effect of public and private investments on the economic growth of Bangladesh. Data from the period between 1972–1973 and 2010–2011 were analysed using Engle–Granger test and neoclassical growth model of Cobb–Douglas production function along with error correction model. The study concluded that there is a close relationship among public investment, private investment and economic growth of Bangladesh both in the long run and in the short run. While comparing the impact, it is proven that private investment has more influence than public investment on economic growth of the country. Khan and Reinhart (1989) suggested that while investing in the economy, if the same amount of investment is carried out by the private sector, the benefits would be greater. The study was conducted to estimate the relative effects of public and private investments of 24 developing countries. Cobb–Douglas growth model was used to analyse cross-country panel data, ranging from 1970 to 1979. The study revealed that public and private investments have a different effect on the economic growth of the countries. Marginal productivity of private investment is proven to be more than that of public investment.

Phetsavong and Ichihashi (2012) investigated the relationship among public investment, FDI and private investment with regard to economic growth of developing countries. Regression analysis was run in a panel data of 15 countries during a 26-year period from 1984 to 2009, and they concluded that private investment is the factor that makes the best contribution towards economic growth. Similar finding, which supported the above viewpoint, was made by Kuppusamy et al. (2009). They used the ARDL model for a period from 1992 to 2006 in the Malaysian economy for evaluating the impact of information and communications technology (ICT) investment by public and private sectors to economic growth. Despite the heavy investment s made by the public sector, such investments proved to be insignificant in the long run and short run of the model. The ICT investment (apart from the agricultural sector) made by the private sector has benefitted the nation. Johansen co-integration method was used to study the role of investment in the economic growth of Pakistan from 1973 to 2008 in the work of Hashmi et al. (2012). Using the VECM model along with OLS regression, they found that although public and private investments have a significant role in economic growth in the long run, it is only the private investment that has an impact in the short run.

A majority of studies in developing countries came up with a conclusion that both public and private investments have an impact on the economic growth of a country. The study conducted by Beddies (1999) is one among them. He determined the macroeconomic variables that led to the economic growth of Gambia. Using production function approach, data from the 1964–1998 period were analysed and the role played by public investment was considered important in the output generation of the economy. Private capital accumulation through private investment also has a greater influence on boosting the output of the country. In addition to that, the study also recommends the implementation of policies that foster public and private capital accumulation in the Gambian economy. In China, Sahoo et al. (2010) studied about the role played by the macroeconomic factors in the economic development of China. The study was conducted by collecting data from 1975 to 2007, using ARDL along with generalised methods of moments (GMM). The study suggests that public and private investments, labour force and infrastructural development played a major role in the economic development of China. The empirical findings also suggested a bidirectional causality between public and private investments and economic growth of China. The study also highlighted the strong impact of infrastructure development on its economic growth.

A comparison on the effect of public investment on private investment in 19 developing countries with that of 12 developed countries was carried out by Erden and Holcombe (2005) during 1980–1997. They used pooled OLS and generalised least square (GLS), and the findings reveal that in developing countries, public investment complements private investment in contrast with that of developed countries. The determinants of private investment in developing countries are also entirely different from that of developed countries. Unavailability of bank credit is regarded as an important constraint in the growth of private investment in developing countries.

When it comes to Indian scenario, only few studies have been conducted on the topic. Noteworthy among them is the study of Mallick (2002). He estimated the long-run determinants of India’s economic growth from 1950 to 1995, using two models—neoclassical endogenous growth model and export-led growth model. They were later analysed using VAR, and determinants of real output and real private investment of India were found. The findings reveal that the real output in India is influenced by public investment, private investment, human capital and real interest rate. Being an important determinant of economic growth, public investment also influenced India’s private investment. The study highlighted the role of private investment in empowering economic growth by proving that the growth in India is not only supply oriented but also demand oriented. An exact enquiry into the relative effect of public investment and private investment, especially in infrastructure sector, has not been conducted in the Indian scenario. This study aims to fulfil that research gap existing in the related literature.

3. Rationale

Advancement in econometric techniques and reliable data sources leads to pristine accuracy in measuring the impact that public and private investments have on the economic growth of a country. Majority of earlier studies made use of Cobb–Douglas model specification in determining the relationship that investment had on economic growth. Seldom do studies concentrate on modern time series analysis like Johansen’s co-integration technique and VECM in determining the above-mentioned relationship. Moreover, empirical work on infrastructure, which bifurcates public and private investments, is rarely found in India. In every sense, that is, whether in the usage of modern data analysis techniques or in the area in which the study delves into, it stands out from existing studies. This study will not only add to the existing literature but also provide a scope for further research in evaluating the relative impact of public and private investments in other different sectors like agriculture, industry and services in India or other countries.

4. Empirical Strategy

Augmented Dickey–Fuller (ADF) test was used to check the stationarity of the time series data, and VAR was used in determining optimal lag length for the model. Later, Johansen’s co-integration methodology was used to test the long-run relationship between the variables. A VECM was also conducted to identify the speed of error correction of the model. Residual diagnostic test like The Breusch–Godfrey LM test and White test were performed to check the adequacy of the model. Along with these tests, an impulse response function and variance decomposition analysis were conducted to measure the impact of public and private infrastructure investments on the GDP of India. EViews 9.0 was used for the analysis.

The data were obtained from National Account Statistics published by Central Statistical Organisation (CSO) under the Ministry of Statistics and Programme Implementation and Database of World Bank. India’s GDP growth rate was readily available from the World Bank database. The methodology for data mining for the present study is inspired from a series of papers written by Murty and Soumya (2006). Capital formation in industry groups 4, 5 and 7, namely electricity, gas and water supply; construction; and transport, storage and communication has been considered as infrastructure investment for the present study. Yearly data on aggregate infrastructure investment and public infrastructure investment at current prices commencing from 1961–1962 to 2016–2017 were collected. Data were used up to the latest update available. The variables used in the present study were measured in growth rates. From the available data, infrastructure investment growth rate of public and private sectors was segregated. Data for 56 years were sufficient to analyse the impact of public and private infrastructure investments because a structural break of liberalisation was considered while studying the impact of private investment. In order to accommodate this structural break in the study, a dummy variable was introduced which stood for the period prior to liberalisation and period after liberalisation. In this framework, the GDP can be fairly explained as a function of public and private infrastructure investments and thus

GDP = f (PUB, PVT, DV)

where

GDP stands for Gross domestic product,

f used as notation for function of,

PUB represents public infrastructure investment,

PVT represents private infrastructure investment and

DV represents dummy variable.

5. Results and Discussion

Up to the liberalisation period, that is, before the early 1990s, we can see the dominance of public infrastructure investment, maintained around 70–80 per cent of total infrastructure investment. At that time, the contribution of private infrastructure investment was too low. But almost after the beginning of the liberalisation era, we can see that the phenomenon until then reversed and a gradual growth of private infrastructure investment was visible. This trend continued for almost 20 years, and the growth of private infrastructure investment soon after that began to show a negative trend. The positive effect of liberalisation over private infrastructure investment could not prolong for a long time and began to disappear gradually. When it comes to sector-wise analysis, in the electricity, gas and water supply sector, almost 86 per cent of investment is made by the public sector and the rest 14 per cent is by the private sector. In India, electricity, gas and water supply are considered to be a forte of the public sector. In addition to that, private sector faces many restrictions imposed by the government to invest in this sector. But the case is entirely different in the construction industry where a lion’s share of investments (about 82%) made by the private and public sectors contributes to only 18 per cent of investment. The transport, storage and communication sectors are also dominated by the private sector, and it is that sector in which the contribution of the private sector is the highest. Till now, the private sector has invested around ₹18,576.85 billion in this sector. This sector has also attracted over 937 public–private partnership (PPP) projects and thus becomes the most attractive infrastructure sector for the private sector to invest in. This finding is similar to the finding of Neeraj (2013).

When it comes to volatility, in comparison to the public infrastructure investment, the private infrastructure investment shows more volatility. This is due to the fact that public infrastructure investment, to a large extent, is determined according to the discretion of the central government. And while allocating necessary funds to finance infrastructure, the government usually follows a consistent trend.

But this is not the case for the private infrastructure investment. Investment by the private sector is very sensitive to macroeconomic conditions. This sensitivity results in the volatility of private infrastructure investment’s growth rate. Private finance will bloom only when there is a suitable environment for the private parties to invest. If such a situation does not persist, they are not ready to invest. This is the main reason why private infrastructure investment fluctuated more than that of public infrastructure investment.

5.1 Unit Root Test: Augmented Dickey–Fuller

For every time series analysis, checking stationarity is of utmost importance. In the absence of stationarity check, the model will bring spurious results (Havi et al., 2013). Augmented Dickey–Fuller test is a widely used stationarity test to conduct unit root analysis of time series data (Wassell & Saunders, 2000). Table 1 shows the descriptive statistics of four variables used in the study.

Table 1.

Descriptive Statistics

Variable	Mean	Minimum	Maximum	Std. Deviation	Kurtosis	Skewness	Observations
PVT	0.180046	−0.25614	1.102804	0.244558	5.056364	0.932378	56
PUB	0.13516	−0.02453	0.36273	0.087972	2.896544	0.642468	56
GDP	5.346606	−5.23818	10.25996	3.110435	4.423208	−1.03202	56
DV	0.464286	0	1	0.503236	1.020513	0.143223	56

Source: EViews output.

Table 2.

ADF Test Results

Variable

Level

First Differencing

None

Intercept

Trend and Intercept

None

Intercept

Trend and Intercept

Order of Integration

t-Stat

p-Value

t-Stat

p-Value

t-Stat

p-Value

t-Stat

p-Value

t-Stat

p-Value

t-Stat

p-Value

GDP

1.809

0.982

−0.068

0.947

−4.432

0.005

−5.702

0.000

−4.333

0.001

−4.311

0.007

I(1)

PUB

−0.966

0.294

−6.196

0.000

−6.117

0.000

−6.371

0.000

−6.302

0.000

−6.229

0.000

I(1)

PVT

−1.580

0.107

−8.726

0.000

−8.635

0.000

−8.419

0.000

−8.349

0.000

−8.272

0.000

I(1)

0.678

−0.9114

0.777

−2.061

0.5554

−7.280

−7.348

−7.278

I(1)

Source: EViews output.

Table 3.

Optimal Lag Selection

Lag	LogL	LR	FPE	AIC	SC	HQ
0	−100.3101	NA	0.000702	4.090592	4.242108	4.148491
1	−25.75553	134.4906*	7.09e-05*	1.794334*	2.551913*	2.083828*
2	−16.15267	15.81647	9.22e-05	2.045203	3.408844	2.566291
3	−9.039252	10.60039	0.000135	2.393696	4.363401	3.146379
4	4.644768	18.24536	0.000157	2.484519	5.060286	3.468796

Source: EViews output.

Note: * Indicates lag order selected by the criterion.

This test is basically concerned with the estimation of B₁. The null hypothesis (H0: B₁ = 0) is rejected when the p-value is less than 5 per cent level of significance and thus enunciates the variable to be stationary. If it is not rejected at this level (with or without trend and intercept), the same would be checked at the first-level differencing. The Table 2 shows the ADF test results of the four variables under study. The p-value of the variables; GDP, PUB, PVT and DV showed the presence of unit root in their levels. So, first differencing of the variable is done, and the results then showed that the variables became stationary at their first differencing. So, all the variables are integrated at same order, that is, I(1).

5.2 Selection of Optimal Lag Length

Since all the variables are integrated at the same order, that is, I(1), Johansen co-integration method can be applied to check the long-run relationship between the variables. The time series analysis is always sensitive to the selection of optimum lag length to bring valid results. VAR was used to identify the optimum lag length for the model. From the results, all criteria unanimously select 1 as optimum lag length for the model. Table 3 shows different criterions used for optimal lag selection of the model.

LR: Sequential modified LR test statistic (each test at 5% level), FPE: final prediction error, AIC: Akaike information criterion, SC: Schwarz information criterion and HQ: Hannan–Quinn information criterion.

5.3 Johansen’s Co-integration Test

Johansen’s maximum likelihood method of co-integration was developed by Johansen (1988) and has been used in the present study, in order to identify the nature of long-run relationship between the variables. If the variables are integrated at same order—I(1)—and their residuals are integrated at level I(0), it shows that a long-run relationship exists between the variables. For estimating the relationship, Johansen co-integration method is considered superior to other methods (Gonzalo, 1994). The long-run relationship between the variables is determined by the number of cointegrating equations. For estimating the number of cointegrating equations, Trace test and max eigenvalue test are used at 5 per cent level of significance. Table 4 shows the result of Johansen’s co-integration test of the model.

Table 4.

Johansen’s Co-integration Test

Hypothesised Number of CE(s)	Trace Statistic (λ_trace)	5% Critical Value	p-Value	Max Eigenvalue Statistic (λ_max)	5% Critical Value	p-Value
None	91.54709	47.85613	0.0000	38.14853	27.58434	0.0015
At most 1*	53.39856	29.79707	0.0000	34.32442	21.13162	0.0004
At most 2*	19.07414	15.49471	0.0138	18.49753	14.26460	0.0101
At most 3	0.576613	3.841466	0.4476	0.576613	3.841466	0.4476

Source: EViews output.

Notes: * Denotes rejection of the hypothesis at the 0.05 level.

Trace test indicates 3 cointegrating equation(s) at the 0.05 level.

Max eigenvalue test indicates 3 cointegrating equation(s) at the 0.05 level.

Table 5.

Normalised Cointegrating Coefficients

Variables	Coefficients	Standard Errors
PUB	−9.565971	6.95735
PVT	−16.38545	2.86294
DV	−2.288618	0.90694

Source: EViews output.

Table 6.

Testing for the Stationarity of the Residuals-ADF Test

Variable	Level
	None		Intercept		Trend and Intercept		Order of Integration
	t-Stat	p-Value	t-Stat	p-Value	t-Stat	p-Value
Residual	−7.61	0.000	−7.53	0.000	−7.61	0.000	I(0)

Source: EViews output.

Using 1 as optimum lag length, Johansen’s cointegration test was carried out. The null hypothesis of having none, at most 1 and at most 2 cointegrating equations was rejected against the alternate hypothesis in both Trace test statistic and max eigenvalue tests. Hence, we can conclude that under this model, there are three cointegrating equations.

Table 5 shows the normalised cointegrating coefficients of the Johansen’s co-integration test. All the three independent variables, including the dummy variable, have a long-term impact on the GDP. The signs of normalised cointegrating coefficients are reversed when it comes to interpretation. It means that there exists a positive relationship between the variables so that when PUB, PVT and DV increase, GDP is also seen to be on the rise. So, the three variables have a positive relationship with GDP. Moreover, the coefficient of private infrastructure investment shows higher value than that of public infrastructure investment.

The linear function of the variables, namely GDP, PUB, PVT and DV, was hence proved as I(0), and this ensures that the above variables are cointegrated.

Table 7.

Estimates of Error Correction Term c(1)

Coefficient	−1.172042
Std. Error	0.200030
t-Statistic	−5.859341
Prob	0.0000

Source: EViews output.

Table 8.

Residual Diagnostics Test Results

Diagnostic test	Results	Hypothesis	Decision
LM Test	p-Value: 0.9247LM-stat: 8.714	H₀: There is no autocorrelation among the residuals. H₁: There is autocorrelation among the residuals.	p-Value > 0.05 H₀ is accepted.
White Test	p-Value: 0.0823Chi-square: 313.469		p-Value > 0.05H₀ is accepted.

Source: EViews output.

5.4 Vector Error Correction Model

When the variables are integrated at same order, it indicates co-integration between the variables in the long run. If the residuals from the model are found to be stationary, there is a need for VECM. The ADF test results of the residual are presented in Table 6. The results clearly state that the residuals are integrated at level, hence I(0). So, VECM is applied in the model to estimate the error correction coefficient and speed of adjustment. The results are presented in Table 7.

c(1) is the coefficient of the speed of adjustment of the long run in a VECM. The error correction is said to be consistent when its coefficient is negative and is statistically significant. Here, the error correction value of the VECM is −1.17 and statistically significant at the 5 per cent level of significance. When the error correction term lies between −1 and −2, Narayan and Smyth (2006) state that rather than converging to the error correction process directly, it keeps on fluctuating in the long run in a decreasing manner. When this process is completed, the convergence to the equilibrium becomes quicker and finally it comes to an equilibrium state.

5.5 Residual Diagnostics Analysis

It is strongly recommended to conduct a residual diagnosis for testing model adequacy in every time series analysis. It is only when a model is free from heteroscedasticity and autocorrelation problems, the findings from the same is considered to be valid. Table 8 shows the variance decomposition analysis of GDP in India.

The Breusch–Godfrey LM test is a test for checking the problem of serial correlation in a model. As the p-value is higher than the significance level of 5 per cent, we fail to reject the null hypothesis of having no autocorrelation in the residuals. Hence, the model has been proved to be free from the problem of autocorrelation. Similar is the case with heteroscedasticity check. White test is performed in order to check whether variance of errors in a model is equally distributed or not. As the p-value is greater than 5 percent significance level, we fail to reject the null hypothesis of having homoscedasticity in variance. Thus, the model is free from the problem of heteroscedasticity.

5.6 Impulse Response Function

Further, the impulse response function of GDP to public infrastructure investment and private infrastructure investment for a period of 10 years is also observed. Impulse response analysis can be used as a method to monitor the response of the dependent variable over a time period when a shock is given to the independent variables in a model.

Figure 1 shows that after giving a shock in public and private infrastructure investments, the GDP is rising as an initial response. When we compare the quantum of rise in both scenarios, the response of GDP to private infrastructure investment is more active when compared to public infrastructure investment. Both fall after the rise, but private infrastructure investment again rises where public infrastructure investment is stagnant. The impact of shock almost disappears from the seventh period onwards and becomes stable gradually. From this, it is clear that although public and private infrastructure investments are having an impact on GDP, the impact of private infrastructure investment on GDP is more strong and noticeable.

Figure 1.

Source: EViews output.

Table 9.

Variance Decomposition of GDP

Period	PUB	PVT
1	0.000000	0.000000
2	0.240217	4.827692
3	1.653147	6.551103
4	1.599720	6.360024
5	1.724224	6.437890
6	1.735699	6.373483
7	1.721516	6.303182
8	1.705869	6.228745
9	1.695462	6.155263
10	1.688287	6.087007

Source: EViews output.

5.7 Variance Decomposition Analysis

Variance decomposition analysis estimates the pass-through of external shocks to the variables in a model. It can be used to access the proportion of the movement of dependent variables due to their own shock and due to the shock of other independent variables over the sample period.

Table 9 shows the percentage of forecast error variance of dependent variable, here GDP, has been explained by independent variables, public infrastructure investment and private infrastructure investment. When the influence of public infrastructure investment and private infrastructure investment are compared, the latter seems to be more explainable for variances in GDP. Over the forecasted period, public infrastructure investment accounts for only around 1 per cent change in GDP, whereas private infrastructure investment stands around 6 per cent. At the 10th year, we can see that only 1.68 per cent of change in variance is caused by public infrastructure investment, whereas private infrastructure investment becomes the cause of around 6.08 per cent of change in GDP. This shows the strong long-run impact and the ability of private infrastructure investment to influence the variations in the GDP in the given model.

6. Conclusion and Implications

Although there is an outspread opinion that private investment has more productivity than public investment, this hypothesis has not been systematically tested in any of the earlier studies in India. In the absence of solid empirical evidence, it is impractical to assume it to be true. This study mainly focused on bringing persuasive empirical evidence in this concern. Based on the empirical evidence discussed earlier, it is evident that both public infrastructure investment and private infrastructure investment are having significant impact on the economic growth of India. Moreover, the relative contribution of public investment and private investment was also measured and findings which came up in this study are in line with majority findings of past literature that, when compared with public investment, it is private investment that is capable of giving a better impetus to economic growth in India. Therefore, one could say that the hypothesis that states that private investment must be favoured to improve the development of the economy has got some empirical support in the infrastructure sector in India. It is also important to note that if private capital is affected by public capital, then the identification of that separate spending portion of government is very important.

These findings can act as an ally to government in infrastructure planning process. This becomes even more prominent in the present scenario wherein the private investment has started showing a sluggish trend. Private investment being of paramount importance to the economic growth of our nation, there is an urgent need to revive this downward trend in private infrastructure investment. The advent of new ideas like PPP and its success stories all over the world are something for us to look upon. The private participation in infrastructure should be shown the green light, and our government should make necessary facilities so as to provide the much needed fillip to private investment. The NDA government’s ‘Make in India’ programme in 2014 was aimed to bolster the private investment, especially in the manufacturing sector, but it is still in the budding stage itself. Though private infrastructure investment in India was improved after the liberalisation programme, the statistics in recent years report a downward trend. This sagging growth rate of private infrastructure investment will have its bearing not only on the GDP of our country but also on the other sectors as well. Even at the time of budget preparation, the government is making an effort to propel private infrastructure investment.

According to CRISIL, the infrastructure investment in India would be on the rise in the coming years. There would soon be a situation where public investment will not be able to handle the growing infrastructure needs. Comprehensive retooling of public–private partnership frameworks and deepening of the infrastructure financing ecosystem is a solution given by Ashu Suyash, CRISIL MD and CEO. Latest statistics show that the share of private investment in infrastructure compared with that of public infrastructure investment has fallen from 37 per cent in 2008 to 25 per cent in 2018. So, in order to increase the private infrastructure investment, the government should develop corporate bond market instruments which will pave the way for more private investment that would in turn pull up the economy. Innovative financial instruments in infrastructure sector like ‘Masala Bonds’, recently issued by Kerala Infrastructure Investment Fund Board (KIIFB), was a noteworthy movement that occurred in India, and it was also said that other states were also impressed and approached the state to know how to get it. Further, government is seeking PPPs, other private mode of investment in infrastructure and even showing open doors to foreign investment so as to increase private participation in infrastructure. The government of India has made FDI regime in infrastructure so liberal. It has allowed 100 per cent FDI in almost all the sectors of infrastructure like telecom and electricity generation, which can be processed through automatic route. But in the actual practice, India has an account of around 22 per cent of FDI in infrastructure sector of India. Further, PPPs also play a major role in reducing the funding deficit to meet the increasing demand of infrastructure in sectors such as water, energy and transport. A developing country like India, where infrastructure needs are numerous and diverse, PPPs help in filling gaps caused by the scarcity of public funds. But the growth of PPPs in India is still in budding stage. In spite of all the efforts taken by the government, investment in infrastructure is still inadequate, and it shows that authorities must go back to the drawing board and must bring innovative ideas and policies to boost both public and private infrastructure investments in India.

Footnotes

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.

Funding

The first author is receiving JRF scholarship from UGC for her doctoral research in infrastructure investment (in general).

References

Afonso

St. Aubyn

(2008). Macroeconomic rates of return of public and private investment: crowding-in and crowding-out effects (Working Paper, No. 864). European Central Bank.

Aschauer

D. A.

(1989). Is public expenditure productive. Journal of Monetary Economics , 23(2), 177–200.

Aziz

(2014, February 26). Why do Keynesians prefer government spending over private spending? The Week . https://theweek.com/articles/450250/why-keynesians-prefer-government-spending-over-private-spending#:~:text=When%20the%20economy%20is%20booming,chance%20of%20the%20economy%20overheating.&text=First%2C%20government%20spending%20can%20reduce%20unemployment%20by%20directly%20creating%20jobs

Beddies

C. H.

(1999). Investment, capital accumulation, and growth: Some evidence from the Gambia 1964–98 (Working Paper, WP/99/117). IMF.

Bédia

(2007). Relative effects of public and private investment on Cote d’lvoire’s economic performance. Applied Econometrics and International Development , 7(1), 151–158.

Belloc

Vertova

(2004). How does public investment affect economic growth in HIPC? An empirical assessment (WP Number 416). Università degliStudi di Siena, Departimento Di Economia Politica.

Coutinho

R. M.

Gallo

G. M.

(1991). Do public and private investment stand in each other’s way? (WDR Background Paper). World Bank.

Crowder

W. J.

Himarios

(1997). Balanced growth and public capital: An empirical analysis. Applied Economics , 29(8), 1045–1053.

Devarajan

Vinaya

Zou

H.-F.

(1996). The composition of public expenditure and economic growth. Journal of Monetary Economics , 37(2), 313–344.

10.

Epaphra

Massawe

J. A.

(2016). Investment and economic growth: An empirical analysis for Tanzania. Preprints , 2016, 1–46.

11.

Erden

Holcombe

R. G.

(2005). The effects of public investment on private investment in developing economies. Public Finance Review , 33(5), 575–602.

12.

Erenburg

S. J.

(1993). The real effects of public investment on private investment. Applied Economics , 25, 831–837.

13.

Evans

Karras

(1994). Is government capital productive? Evidence from a panel of seven countries. Journal of Macroeconomics , 16(2), 1–11.

14.

Ghali

K. H.

(1998). Public investment and private capital formation in a vector error correction model of growth. Applied Economics , 30(6), 837–844.

15.

Ghura

(1997). Private investment and endogenous growth: Evidence from Cameroon (Working Paper, WP/97/165). International Monetary Fund (IMF).

16.

Gonzalo

(1994). Five alternative methods of estimating long-run equilibrium relationships. Journal of Econometrics , 60(1–2), 203–233.

17.

Haque

S. T.

(2013). Effect of public and private investment on economic growth in Bangladesh: An econometric analysis. Research Study Series No FDRS05/2013, Finance Division, Ministry of Finance, Bangladesh , 2013, 104–126.

18.

Hashmi

M. H.

Akram

Hashmi

A. A.

(2012). Role of investment in the course of economic growth in Pakistan. International Journal of Academic Research in Economics and Management Sciences , 1(5), 48–61.

19.

Havi

E. D.

Enu

Osei Gyimah

Attah-Obeng

Opoku

(2013). Macroeconomic determinants of economic growth in Ghana: Cointegration approach. European Scientific Journal , 9(19), 156–175.

20.

Johansen

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

21.

Khan

M. S.

Kumar

M. S.

(1997). Public and private investment and the growth process in developing countries. Oxford Bulletin of Economics and Statistics , 59(1), 69–88.

22.

Khan

M. S.

Reinhart

C. M.

(1989). Private investment and economic growth in developing countries (Working Paper WP/89/60). IMF.

23.

Kirkpatrick

Parker

Zhang

Y.-F.

(2006). Foreign direct investment in infrastructure in developing countries: Does regulation make a difference. Transnational Corporations , 15(1), 143.

24.

Kuppusamy

Raman

Lee

(2009). Whose ICT investment matters to economic growth: Private or public? The Malaysian perspective. The Electronic Journal of Information Systems in Developing Countries , 37(1), 1–19.

25.

Lighthart

J. E.

(2000). Public capital and output growth in Portugal: An empirical analysis (Working Paper WP/00/11). IMF.

26.

Lynde

Richmond

(1992). The role of public capital in production. The Review of Economics and Statistics , 74(1), 37–44.

27.

Lynde

Richmond

(1993). Public capital and total factor productivity. International Economic Review , 34(2), 401–414.

28.

Mallick

S. K.

(2002). Determinants of long-term growth in India: A Keynesian approach. Progress in Development Studies , 2(4), 306–324.

29.

Munnell

A. H.

(1990). Why has productivity growth declined? Productivity and public investment (pp. 3–22). New England Economic Review.

30.

Munnell

A. H.

(1990a). Is there a shartfall in public capital investment? An overview (pp. 1–20). Federal Reserve Bank of Boston.

31.

Murty

K. N.

Soumya

(2006). Effects of public investment in infrastructure on growth and poverty in India (Working Papers 2006-006, pp. 1–41). Indira Gandhi Institute of Development Research.

32.

Narayan

P. K.

Smyth

(2006). What determines migration flows from low-income to highincome countries? An empirical investigation of Fiji-U.S. Contemporary Economic Policy , 24(2), 332–342.

33.

Nataraj

(2014). Infrastructure challenges in India: The role of public-private partnerships (Occasional Paper #49). ORF.

34.

Nazmi

Ramirez

M. D.

(1997). Public and private investment and economic growth in Mexico. Contemporary Economic Policy , 1997, 65–75.

35.

Neeraj

(2013). The Indian infrastructure sector: Investments, growth, and prospects . India Brand Equity Foundation.

36.

Pereira

A. M.

de Frutos

R. F.

(1999). Public capital accumulation and private sector performance. Journal of Urban Economics , 46(2), 300–322.

37.

Phetsavong

Ichihashi

(2012). The impact of public and private investment on economic growth: Evidence from developing Asian countries (DEC Discussion Paper). Hiroshima University.

38.

Ram

Ramsey

D. D.

(1989). Government capital and private output in the United States. Economics Letters , 30(3), 223–226.

39.

Ramirez

M. D.

(1996). Public and private investment in Mexico and Chile: An empirical test of the complementarity hypothesis. Atlantic Economic Journal , 24(4), 301–320.

40.

Sahoo

Ranjan Kumar

Geethanjali

(2010). Infrastructure development and economic growth in China (Discussion Paper No. 261). Institute of Developing Economies.

41.

Serven

Solimano

(1993). Economic adjustment and investment performance in developing countries: The experience of the 1980s. In Serveii

Solirnano

(Eds.), Striving for growth after adjustment: The role of capital formation (pp. 147–179). The World Bank.

42.

Sturm

J. E.

de Haan

(1995). Is public expenditure really productive. Economic Modelling , 12(1), 60–72.

43.

Tatom

J. A.

(1991). Public capital and private sector performance. Review of the Federal Reserve Bank of St Louis , 78(3), 3–15.

44.

Varshney

(2008). Financing of infrastructural development in India since 1991 . Aligarh Muslim University.

45.

Wassell

C. S.

Saunders

P. J.

(2000). Time series evidence on social security and private saving: The issue revisited (Unpublished Manuscript). Department of Economics, Central Washington University.

46.

Zou

(2006). Empirical studies on the relationship between public and private investment and GDP growth. Applied Economics , 38(11), 1259–1270.