The Global Crisis was followed by a massive increase in public sector borrowing. Total outstanding public debt doubled from $35 trillion in 2007 to $70 trillion at the end of 2018. Over the same period, public debt increased from 70% to 102% of GDP in advanced economies and from 35% to 50% of GDP in emerging and developing economies.
This rapid increase in government debt sparked a large literature aimed at estimating the effects of public sector borrowing on economic activity. Following the influential contributions of Reinhart and Rogoff (2010), a large number of papers used country-level data to establish the presence of a negative correlation between government debt and each of economic growth and investment (e.g. Cecchetti et al. 2011, Checherita-Westphal and Rother 2012, Kumar and Woo 2015), but also highlighted the presence of substantial cross-country heterogeneity (Eberhardt and Presbitero, 2015, Kourtellos et al. 2013), and challenged the presence of debt thresholds (Chudik et al. 2017).
Have you read?
While there is strong evidence that debt is negatively correlated with economic growth and investment (Fatás et al. 2019), the cross-country literature has been less successful in establishing the presence of a causal link going from public debt to economic growth and investment (Panizza and Presbitero, 2014).
Reverse causality is a particularly important issue for the study of the link between debt and growth and investment, respectively. Traditional Keynesian policies and neoclassical models of optimal policy suggest that countries should run deficits, and hence accumulate debt, in bad times and surpluses in good times. If shocks to growth are persistent (Cerra and Saxena, 2008), the presence of a countercyclical fiscal policy can generate a long-run negative correlation between debt and growth, where it is low growth that causes high debt and not the other way around.
While economists disagree on the effects of fiscal policy on aggregate economic activity (Perotti, 2008), most macroeconomic models agree in predicting that fiscal deficits, and the subsequent increase in public debt, should reduce private investment when measured as a share of total aggregate demand. This consensus notwithstanding, it is difficult to use cross-country data to move beyond correlations and show that public debt has a causal effect on private investment.
In recent research (Huang et al. 2018), we use data for nearly 550,000 firms in 69 countries over 1998-2014 and provide a direct test for the crowding out effect emphasised by the economic literature.
Following Huang et al.’s (2019) work on the link between local government debt and corporate investment in Chinese cities, we study the behaviour of corporate investment and thus describe a channel through which public debt directly affects economic activity. We show that government debt affects corporate investment by tightening the credit constraints faced by private firms.
Standard models of crowding out focus on the interest rate channel: an increase in government spending puts upward pressure on interest rates which, in turn, leads to lower private investment. The mechanism through which public debt crowds out private investment can also go through quantities instead of prices. In the presence of credit rationing and financial frictions, government debt can be particularly deleterious for firms which have restricted access to credit.
We test this hypothesis by studying whether the crowding out effect of public debt is stronger for firms which are more likely to be credit constrained (i.e. unlisted, small and medium-sized, and young firms).
We start by showing that the negative correlation between investment and public debt which is present in national accounts data is robust to measuring investment with aggregates obtained from firm-level data (Figure 1).
Figure 1 Correlation between public debt and corporate investment
Note: Binned scatterplot (60 bins) controlling for country and year fixed effects.
Source: Own elaboration based on IMF World Economic Outlook
To move beyond correlations and address endogeneity concerns, we first use an empirical strategy that builds on Rajan and Zingales (1998) and show that high levels of government debt are particularly damaging for industries that, for technological reasons, need more external financial resources. Next, we build on Love (2003) and Huang et al. (2019) and use firm-level data to show that the sensitivity of investment to internal funds increases with the level of government debt.
Figure 2 shows that when public debt is at 25% of GDP, the correlation between investment and cash-flow is just above 9%, but this correlation goes well above 10% when public debt surpasses 100% of GDP. This finding is consistent with the idea that higher level of public debt tightens the credit constraint faced by private firms.
Figure 2 Correlation between investment and cash flow at different levels of public debt
As our empirical approach focuses on within-country-year variation, it controls for all macroeconomic shocks that can jointly affect public debt and investment and rules out any concern of reverse causality linked to fact that governments may decide to run deficits (and accumulate debt) during recessions.
We show that higher levels of public debt increase the correlation between investment and cash flow only for firms that are more likely to be credit constrained – namely unlisted, small, and young firms. However, we find that public debt has no effect on the correlation between cash and investment of listed, well-established, and large firms (i.e. firms that are less likely to face tight credit constraints).
Hence, besides addressing most endogeneity and model specification concerns, and therefore providing evidence of a causal link going from debt to private investment, our empirical exercises allow us to test for the presence of a credit-rationing channel by showing that public debt is particularly damaging for credit-constrained firms.
Two caveats are in order. First, there is a trade-off between our ability to identify the causal effects of debt and that of estimating its aggregate effect on the economy.
While Figure 1 shows that there is a negative correlation between investment and public debt, these are simple correlations that cannot tell us anything about causality. Regressions that control for country-year fixed effects give us a cleaner identification strategy and can provide a test of a specific type of crowding out. However, in these regressions we only observe within-country-year differences and cannot say anything about the effect of debt on total investment because all macroeconomic effects are captured by the fixed effects. Therefore, it could be that higher levels of debt increase investment for all industries and firms, but that investment increases relatively less for firms that are more likely to face credit constraints, as Demirci et al. (2018) suggested as well.
Second, we focus on just one component of aggregate demand. Hence, we cannot say anything on the desirability of the fiscal expansion that followed the Global Crisis and on the desirability of Keynesian policies more generally.
Authors’ note: The views expressed in this paper are those of the author(s) and do not necessarily represent the views of the IMF, its Executive Board, or IMF management.
Cecchetti, S G, M S Mohanty, and F Zampolli (2011), “Achieving Growth Amid Fiscal Imbalances: The Real Effects of Debt”, Proceedings - Economic Policy Symposium - Jackson Hole, Federal Reserve Bank of Kansas City, pp. 145-196.
Cerra, V, and S Chaman Saxena (2008), “Growth dynamics: the myth of economic recovery”, American Economic Review 98(1): 439-57.
Checherita-Westphal, C, and P Rother (2012), “The impact of high government debt on economic growth and its channels: An empirical investigation for the euro area”, European Economic Review 56(7): 1392-1405.
Chudik, A, K Mohaddes, M H Pesaran, and M Raissi (2017), “Is there a debt-threshold e.ect on output growth?”, Review of Economics and Statistics 99(1): 135-150.
Demirci, I and, J Huang and C Sialm (2018), “Government Debt and Corporate Leverage: International Evidence”, Journal of Financial Economics (forthcoming)
Eberhardt, M, and A Presbitero (2015), “Public debt and growth: Heterogeneity and non-linearity”, Journal of International Economics 97(1): 45-58.
Fatás, A, A Ghosh, U Panizza, and A F Presbitero (2019), "The Motives to Borrow," CEPR Discussion Paper 13735.
Huang, Y, M Pagano, and U Panizza (2019), “Local Crowding Out in China,” Journal of Finance (forthcoming)
Huang, Y, U Panizza, and R Varghese (2018), "Does Public Debt Crowd Out Corporate Investment? International Evidence," CEPR Discussion Paper 12931.
Kourtellos, A, T Stengos, and C M Tan (2013), “The effect of public debt on growth in multiple regimes”, Journal of Macroeconomics 38: 35-43.
Kumar, M S, and J Woo (2015), “Public debt and growth”, Economica 82: 705-739.
Love, I (2003), “Financial development and financing constraints: International evidence from the structural investment model”, The Review of Financial Studies 16(3): 765-791.
Panizza, U, and A F Presbiter (2014), “Public debt and economic growth: is there a causal effect?”, Journal of Macroeconomics 41: 21-41.
Perotti, R (2008), “In Search of the Transmission Mechanism of Fiscal Policy,” in D Acemoglu, K Rogoff, and M Woodford (eds), NBER Macroeconomics Annual, Volume 22: 169 – 226, Chicago University Press
Rajan, R, and L Zingales (1998), “Financial Dependence and Growth”, American Economic Review 88(3): 559-86.
Reinhart, C M, and K S Rogoff (2010), “Growth in a Time of Debt”, American Economic Review P&P 100(2): 573-78 (see also a Vox column).