Debt-financed public investment in developing countries: Does the efficiency of public investment matter?

This study examines whether government spending efficiency is associated with differential effects of public investment on debt-to-GDP ratio for a panel data consisting of 16 developing countries in Asia-Pacific region over the period 2007-2017. Public investment is central to implementing the UN 2030 Agenda for Sustainable Development — but high debt-to-GDP ratio poses a key risk. The empirical results indicate that public investment efficiency moderates debt-to-GDP ratio whereas public investment in the midst of public sector corruption accentuates debt-to-GDP ratio. The results have important policy implications.

As demonstrated in Abiad et al. (2017) and Delong and Summers (2012); in the short term, an increase in public investment as a share of potential GDP (i) (leads to a change in the debt-to-potential GDP ratio () given by: In which  is the fiscal multiplier and  is the marginal tax rate The efficiency of public investment is central to determining the size of the fiscal multiplier and the elasticity of revenue with respect to output.
Inefficiencies in the public investment process, such as poor project selection, implementation, and monitoring, can result in a fraction of public investment turning into productive infrastructure, undermining the long-term output gains (Pritchett 2000).
Public investment efficiency contributes to higher output by increasing the stock of capital. The extent to which increases in public capital can raise output is a key factor in determining the sequence of public debt-to-GDP ratio. Over time, the increase in public capital will affect the debt-to-GDP ratio by affecting annual debt-financing burden, which is equal to the difference between the real government borrowing rate (r) and the GDP growth rate (g) multiply by the initial change in the debt-to-GDP ratio: (rg)  = (rg) (1 -)i (2) How the financing burden will affect the debt-to-GDP ratio in the long term depends on the parameters of equation (2) and also the elasticity of output to public capital, . In the long term, an increase in public investment may lead to an increase in output (Y), which will generate long-term future revenues: Where  is the long-term elasticity of output to public capital and is the initial output-to public capital ratio. Equations (2) and (3) jointly imply that if the returns to public capital (short-term multipliers and the elasticity of output to public capital) are large enough, such that: Then an increase in public investment will be self-financing.

Related Literature and Studies
A wide range of empirical literature has emerged over the last two decades showing that the quality of institutions matter for development. Such findings elevated governance as a determinant of economic development. Studies that are directly related to public debt are those that consider the effect of corruption on debt. Kaufmann (2010) points to a strong correlation between corruption and fiscal deficits in industrialized countries; it suggests that if Greece's levels of corruption was the same to Spain, its budget deficit over the last five years would be 2.5% of GDP rather than 6.5%. Grechyna (2012) builds a model that relates the level of government debt to the degree of corruptness of the public officials in developed economies and finds that public corruption results in higher public debt levels. Gonzalez-Fernandez and Gonzalez-Velasco (2014) used panel data to analyzed the relationship between the shadow economy and corruption as determinants of public debt and their results find that corruption has positive and significant impacts on regional public debt in Spain.
Debt burdens are of great concern to the developed countries but the developing country debt crises is of greater concern and a recurrent phenomenon. Less focus has been paid to possible interactions between government spending efficiency, public investment and public debt. This is important because public investment will continue to be fundamental in financing development goals which has the tendency to raise debt ratios. Therefore, addressing spending inefficiencies is critical-countries need to spend not only more, but better. We conjecture that countries could make significant savings through efficiency efforts.
In several countries, increased by public investment does not lead to productive capital (Pritchett 1996). A significant proportion of the expected returns from spending on health, education, and infrastructure is lost due to spending inefficiencies. In the area of health, Grigoli and Kapsoli (2018) find that countries with low efficiency index could raise healthy life expectancy by up to five years by addressing inefficiencies. In the area of education, Grigoli (2015) finds that addressing inefficiencies could help increase enrollment by more 30 percentage points in developing countries. In the area of infrastructure, IMF (2015) finds that more than 30 percent of investment is lost through inefficiency with larger losses in developing countries. In addition, cross-country regressions by (IMF 2015) suggests that the quality of institutions is the main determinant of public investment efficiency and the efficiency scores are a function of a set of explanatory variables such as: the quality of institutions, measured by control of corruption and regulatory quality. Overall, the estimates show a positive relationship between public investment efficiency and the quality of institutions.
Efficiency refers to the case where public goods and services are provided at the minimum cost. High levels of corruption, for example, may be a cause of public investment inefficiency. This research study is novel attempt to examine the effect of public investment efficiency on debt-to-GDP ratio in selected developing countries in Asia-Pacific countries.

Debt Composition
Developing countries mobilize part of their resources by borrowing from internal and external sources to finance their development activities. These sources gradually build up the debt stock of the country. Such debt stock demands regular debt servicing, that is, principal and interest payment, which consumes scarce resources that can be used for financing development. Excessive borrowing to finance deficits drains the resources of the developing countries through higher cost of servicing debts. Figure 1 shows that in least developed countries (LDCs): multilateral debt increased marginally on aggregate; some countries increased their borrowing while others reduced their borrowing.

Figure 1
Source  million to 1 billion. Domestic debt increased in almost all the countries; Solomon Islands reduced its domestic debt while PNG reported the largest increase with a debt stock of 4.9 billion. Domestic debt now accounts for more than 50% of the total public debt.
On average, multilateral debt seems to be running out of steam as bilateral debt fills the void. However, developing Asia Pacific countries have raised their appetite for commercial debt; not only is debt growing, its structure is changing. The share of commercial-and more costly-debt has increased.

Figure 2
Source: World Bank Development Indicators Figure 2 shows debt-to-GDP ratio varies across countries. Bhutan, India, Mongolia, and Sri Lanka have all surpassed the 60% threshold, while others such as Uzbekistan, Kazakhstan, Indonesia and Nepal have more fiscal space. Countries can have high debt-to GDP ratio and not be in debt distress or at high risk of debt distress. Composition of debt matters. For example, Bhutan has a debt-to-GDP ratio above 100%. Bhutan's interest payment as a percentage of revenue is relatively low at 5% -mainly because its public debt are largely concessional loans. However, interest payments in Sri Lanka is significantly higher at 36% than Bhutan despite similar tax revenue and higher debt-to-GDP ratio. Sri Lanka's debt composition is markedly different; non-concessional loans in 2006 as % of GDP was 7% and increased to 55% in 2018. Concessional loans exert far less pressure on debt service obligations than commercial loans, and it is the major reason (among others ) why interest payment is high in Sri Lanka.

Figure 3
Source: World Bank International Debt Statistics Debt composition indeed matters for debt sustainability. Figure 3 shows China, Turkmenistan and India have the lowest external debt as a percentage of total government debt while Bhutan, Cambodia and Timor Leste have the highest external debt and vice versa. The choice between external and domestic financing is not the focus of this research. However, in low-income countries, highly concessional external debt is usually a better choice to domestic debt in terms of financial risks and costs, even in the face of a probable devaluation, subject to some caveats.  investment. However, it is a good idea to look at public investment in relation to government debt. Armenia's debt to GDP ratio jumped from 21% to 45% and within the same period, public investment shrank from 5% to 3%. Kazakhstan's debt to GDP ratio went from 9% to 20% and within the same period, public investment reduced from 6% to 5%. Sri Lanka's debt to GDP ratio is almost 80% but public investment is 12%. In comparison, Laos has a smaller debt to GDP ratio of 59% but public investment is 28%. The same goes for Vietnam with debt to GDP ratio of 58% but public investment is 25%. Some countries have low debt to GDP ratio and also low public investment for example Indonesia and Kazakhstan which means there is ample room to scale up public investment. Nonetheless, public investment has generally not been scaled up commensurately with the increase in public debt, and where this is significantly the case, public debt went into government consumption, and/or debt repayments.

Data and Estimation Results
The data covers the 2007-2017 period for 16 Asia-Pacific countries which are: Armenia, Azerbaijan, Bangladesh, Bhutan Cambodia, Indonesia, India, Kazakhstan, Malaysia, Mongolia, Nepal, Philippines, Sri Lanka, Thailand, Uzbekistan, and Vietnam. The estimation is carried out using panel fixed effects regression. The countries are chosen due to data availability. Time frame is chosen due to limited data for government spending efficiency measure. The dependent variable is the ratio of public debt to GDP, which is measured as the central government debt, total (% of GDP). Central government debt refers to the debt by the federal government, while general government debt refers to overall debt, including states and municipalities.
The independent variables of interest are the interactions between public investment, government spending efficiency and public sector corruption.
We measure public investment as general government investment (% of GDP). We measure government spending efficiency as how efficient is the government in spending public revenue (1 = extremely inefficient; 7 = extremely efficient ). We measure public sector corruption using the Corruption Perceptions Index (CPI): an index published annually by Transparency International which ranks countries by their perceived levels of public sector corruption, as determined by expert assessments and opinion surveys. The CPI generally defines corruption as the misuse of public power for private benefit. Here the estimate of corruption ranges from 0 (totally corrupt) to 100 (totally not corrupt). The control variables are: GDP per capita growth (annual %), GDP growth may not translate into growth in GDP per capita due to population growth; trade openness, the sum of imports and exports by GDP; inflation, the percentage change in the GDP deflator; and interest payments, (% of revenue). Table 1 shows the regression results. The effect of public investment on debt to GDP ratio is ambiguous; however, it has no significant effect on the ratio of public debt to GDP in both regressions. More importantly, public investment x government efficiency has a statistically significant negative effect on the ratio of public debt to GDP. On the other hand, public investment x public sector corruption has a statistically significant positive effect on the ratio of public debt to GDP. The results imply that the multiplier effect of increased public investment in more corrupt countries is relatively lower and the multiplier effect of increased public investment in countries with more efficient public spending is relatively higher. This is somewhat similar to Abiad et al. (2015) results that found public investment shocks lead to significant medium-term reduction in the debt-to-GDP ratio of countries with high public investment efficiency but increase debt-to-GDP ratio in countries with low public investment efficiency in a sample of developed countries. However, our research focused on a sample of developing countries -which have relatively weak public investment management institutions -and introduced public corruption index as a robustness check. The growth of GDP per capita as expected has a statistically significant negative effect on the ratio of public debt to GDP for both regressions.
Higher growth lowers the overall debt-to-GDP percentage. Conversely, declines in the long-run economic growth rate drive increases in the debt-to-GDP ratios. The higher the inflation, the lower the ratio of public debt to GDP; inflation can rise and decrease the real value of the domestic debt, since debt is denominated in its own currency; however, inflation can also rise and increase the real value of the external debt through exchange rate depreciation. The net effect depends on the debt composition. The higher the interest payments, the higher is the ratio of public debt to GDP; higher interest payments increase the cost of refinancing the stock of existing debt. The more open an economy becomes, the higher the ratio of public debt because changes in a country's export and import prices affects the decision to issue new debt. In addition, capital account openness facilitates capital inflows which results in higher debt levels. The regression estimates should be interpreted with caution due to the sample size. However, the central message remains the same: public investment efficiency indeed matters in moderating or accentuating public debt ratios.

Conclusions
This study examined the effect of public investment efficiency on debt ratios. The empirical results show that public investment efficiency has a statistically significant negative influence on public debt which implies that public investment efficiency moderates the positive effect of public investment on debt to GDP ratio. However, when public investment interacts with public sector corruption, the variable becomes statistically significant and positive which supports the claim that public investment in the midst of corruption worsens debt ratios. The research study has important policy implications: improving actions against public sector corruption or raising the efficiency of public investment could help moderate debt ratios. No. of countries 16 Note: Robust standard errors reported in parenthesis. ***, **, *. Denotes significance level at 1%, 5%, 10%, respectively