THE IMPACT OF PUBLIC DEBT ON ECONOMIC GROWTH.

• Author: Salmon, Jack

Following the 2007-2008 global financial crisis (GFC) and subsequent sovereign debt crisis in Europe, there has been a renewed interest in exploring the relationship between government debt and economic growth. One of the cornerstone studies on the subject that triggered an emergence of new literature was Carmen Reinhart and Kenneth Rogoffs "Growth in a Time of Debt" (2010), which became widely cited and influential among commentators, academics, and policymakers in the debate surrounding austerity and fiscal policy in debt-burdened economies. Much of the research that followed the GFC uses panel data analysis of the debt-growth nexus using datasets from the World Bank, International Monetary Fund (IMF), European Commission, and Organisation for Economic Co-operation and Development (OECD).

A notable pattern emerges from that research: high levels of public debt have a negative impact on economic growth.

The main objective of this survey is to review the existing economic literature published during the period 2010 to 2020 on the relationship between public debt levels and economic growth. In addition, the survey will review the claim that there is a nonlinear debt threshold above which debt has a significant deleterious impact on growth rates.

This article explains how studies were identified for the survey sample, provides an overview of the theories of how public debt impacts economic growth, reviews the findings of the 40 studies in the survey sample, and concludes with some recommendations for future research.

Identification of Study Sample

The survey of literature incorporates datasets on public debt and growth from the World Bank's World Development Indicators, the International Monetary Fund's World Economic Outlook, the European Commission's annual macroeconomic (AMECO) database and Eurostat database, and the OECD's Economic Outlook. Twenty-seven of 40 studies use at least one World Bank database, 14 of 40 use at least one IMF database, 10 of 40 use either AMECO or Eurostat datasets, and 11 of 40 use at least one OECD database.

The choice of studies in the survey is limited to those observing multiple countries (a minimum of seven) to avoid simply observing dynamics that exist in isolated samples or smaller datasets. It was determined that large datasets surveying multiple countries should offer more comprehensive observations of debt-growth dynamics. While an overwhelming majority (33 of 40) of the studies in the survey are published in peer reviewed journals, which may offer higher quality analysis, the survey also includes a smaller number (7 of 40) of nonscholarly articles published by reputable institutions including the Federal Reserve Bank, the Bank for International Settlements, and the World Bank.

For this survey, only articles published in the English language between the years 2010 and 2020 were included. This 10-year period of publications was chosen based on peak interest in the subject following the GFC and the publication of Reinhart and Rogoffs seminal study in 2010. Once these criteria were established, the survey sample was obtained using search terms in Google Scholar such as "debt and growth," "public debt and economic growth," and "debt growth threshold." Some additional studies were added to the survey sample upon reviewing the references and citations of initial studies in the sample. Nineteen of the studies used a sample of advanced and developing nations, 11 used European Union or European countries, 8 used advanced countries only, and just 2 studies focused on developing countries.

Many of the studies in the sample find a negative correlation between debt and growth. More recent studies (e.g., Pegkas, Staikouras, and Tsamadias 2020) rely on Granger causality tests using a vector autoregression model. These tests may be more reliable in determining whether changes in public debt levels are useful in forecasting changes in future growth rates.

The survey studies employ various identification strategies in their empirical analysis, with most studies in the sample using panel data across time, while other studies employ time series analysis, a cross-sectional observation, or a combination of these approaches. A key advantage of using a panel data model approach is that it highlights individual heterogeneity, if there are some differentiating qualities across cross-sections. The larger datasets of the panel data method also allow for a more accurate measurement of the independent effects of the sample, which cross-sectional and time series methods may not account for. In addition, various linear and nonlinear regression methods are adopted across the pool of studies, and many of the studies in the sample employ endogenous or neoclassical growth models.

Theoretical Overview of the Debt-Growth Nexus

While there has been a revival in interest in the relationship between public debt and economic growth since the GFC, economists have long theorized about the various channels through which public debt may affect growth. Since die 1960s, neoclassical economists have noted how increases in taxation, to finance interest payments on die nation's growing domestic and foreign government debt, negatively affect gross capital stock formation (Diamond 1965). Keynesian economists, meanwhile, have argued that rising public debt induces productive public spending and has a positive multiplier effect on the economy (Leao 2013).

The various channels through which high and growing public debt levels adversely affect economic growth include (1) the crowding out of private investment (Elmendorf and Mankiw 1999) as government borrowing competes for funds in the nation's capital markets; (2) higher long-term interest rates caused by an excess supply of government debt and greater credit risk premia (Codogno, Favero, and Missale 2003; Ardagna, Caselli, and Lane 2007; Kumar and Baldacci 2010; Attinasi, Checherita, and Nickel 2011; Von Hagen, Schuknecht, and Wolswijk 2011); (3) higher distortionary taxes to fund future liabilities and rising debt repayments (Dotsey 1994); and (4) an increase in the rate of inflation (Cochrane 2011).

In recent years, New Keynesian economists have argued that debt levels are of little concern as they relate to other economic factors as long as interest rates on the public debt remain below rates of economic growth in the long run (Blanchard 2019). This view ol the debt-growth relationship may overlook existing primary budget deficit dynamics as well as the upward pressures of an increasing debt ratio (public debt as a share of GDP) on long-term interest rates. Acknowledging these uncertainties, more recent observations suggest that large increases in tire debt-to-GDP ratio could lead to much higher taxes, lower future incomes, and intergenerational inequity (Boskin 2020).

Aside from these theoretical arguments, there exists another theory that corroborates the existence of a nonlinear relationship between public debt levels and economic growth-namely, the threshold or nonlinear effect theory. According to this theory, increases in government debt levels have positive growth effects when debt levels are low, but these effects become negative when debt levels increase beyond a certain threshold level (Reinhart and Rogoff 2010).

Figure 1 shows that, at low debt levels, increases in the debt ratio provide positive economic stimulus in line with conventional Keynesian multipliers. Once the debt ratio reaches heightened levels (nonlinear threshold), further increases in the debt level as a percentage of GDP have a negative impact on economic growth (Baum, Checherita-Westphal, and Rother 2013). Existence of a nonlinear threshold would imply that neoclassical theories on the relationship between debt and growth may be well grounded. Such theories suggest that the distortionary impact of future tax increases to achieve debt sustainability will likely lower potential economic output (Barro 1979). Additionally, a nonlinear threshold could suggest that increased government borrowing competes for funds in the nation's capital markets, which in turn raises interest rates and crowds out private investment, confirming the debt overhang theory.

Survey Sample Results

The survey starts by reviewing the nine studies in the sample that do not seek to find a nonlinear threshold. The remainder of this section will then review the 31 studies in the survey that explore whether there is a nonlinear threshold level at which debt has an adverse impact on economic growth.

Studies Not Exploring a Threshold Level

Starting with studies that focus solely on the debt and growth relationship, Calderon and Fuentes (2013) run time series, cross-country growth regressions to test whether public debt hinders growth and whether economic policy ameliorates this effect. Using a large dataset of 136 countries from 1970 to 2010, the authors find a robust negative relationship between public debt and growth. Interestingly, the quality of institutions, sound domestic policy, and outward-oriented policies can help to reduce this negative effect.

Zouhaier and Fatma (2014), using World Bank data, find that the ratio of total external debt to GDP is statistically significant and negatively affects growth. In particular, they find that an increase in the debt ratio by 10 percentage points will cause real GDP growth to fall by 0.28 percentage points. The sample includes 19 developing countries for the period 1990 to 2011. The study runs a dynamic panel regression (Arellano-Bond estimator) with controls for investment, trade openness, inflation, and other factors.

Siddique, Selvanathan, and Selvanathan (2016) use an autoregressive distributed lag (ARDL) model--with controls for trade, population, and capital formation--to observe whether debt as a proportion of GDP affects growth in 40 indebted countries from 1970 to 2007. The authors find that the debt variable has a negative and statistically significant...