Abstract
Governments issue debt for good and bad reasons. While the good reasons—intertemporal tax-smoothing, fiscal stimulus, and asset management—can explain some of the increases in public debt in recent years, they cannot account for all of the observed changes. Bad reasons for borrowing are driven by political failures associated with intergenerational transfers, strategic manipulation, and common pool problems. These political failures are a major cause of overborrowing though budgetary institutions and fiscal rules can play a role in mitigating governments’ tendencies to overborrow. While it is difficult to establish a clear causal link from high public debt to low output growth, it is likely that some countries pay a price—in terms of lower growth and greater output volatility—for excessive debt accumulation.
I. Introduction
1. The issuance of public debt is an important tool of economic policy. Borrowing can help countries to deal with negative shocks, undertake countercyclical fiscal policy, and finance exceptionally large expenditures, such as public infrastructure investment projects. Many governments, particularly in advanced economies, responded to the Global Financial Crisis with an exceptionally large debt-financed fiscal stimulus. In the United States, for instance, the Obama administration in 2009 approved the American Recovery and Reinvestment Act, a plan of USD 831 billion—about 5.5 percent of GDP—aimed at creating jobs and boost investment. In the aftermath of the crisis, government debt increased sharply from 64 percent of GDP in 2007 to above 100 percent in 2012. In advanced economies the average debt-to-GDP ratio rose from about 60 percent in 2007 to over 90 percent in 2016. Stimulus spending and cyclically-lower revenues also resulted in higher public debt—at various levels of government—in many emerging markets as well. China is a case in point. The government embarked on a massive infrastructure and public investment program, spending more than 6 percent of GDP in discretionary stimulus measures and allowing public debt to increase from 29 percent of GDP in 2007 to 44 percent of GDP by 2016. More recently, the big infrastructure push associated with the China’s Belt and Road Initiative is also contributing to growing public debts and possibly to sustainability risks in some emerging market and developing countries (see Box 1).
Balancing Investment Needs with Debt Sustainability, the Case of Ethiopia
The 2030 Sustainable Development Agenda set ambitious targets for inclusive development, which will require a large scale up of investment over a long period of time. While the private sector should play a key role to mobilize resources, public investment is expected to increase significantly in several countries. At the same time, rising public debt is a source of concern in many developing countries, especially in Africa, where, after the sharp decline of debt levels thanks to the debt relief initiatives of the early 2000s, debt-to-GDP ratios are rising again—driven mostly by large primary deficits and the scaling-up of public investment (IMF 2018a). Rising debts and external imbalances are undermining debt sustainability and could pose a threat for future investment plans and sustained economic growth.
A case in point is Ethiopia. Public investment was above 7 percent of GDP in the 2000s and further accelerated in the last 7 years (public investment was at about 15 percent of GDP between 2014 and 2017). This massive scale up of investment was funded by external concessional and non-concessional financing (including large Chinese investment flows), and partly facilitated by restrained government consumption, financial repression and an overvalued exchange rate (World Bank, 2016). This policy mix allowed the expansion of a substantial physical infrastructure and the development of large projects, like the Grand Ethiopian Renaissance Dam and the railway connecting Addis Ababa with the port of Doraleh in Djibouti, which will significantly reduce trade costs and improve the access to global markets for Ethiopian firms. The dam is estimated to cost almost USD 5 billion—about 5 percent of GDP—and once completed it will be the largest hydroelectric power plant in Africa, supplying energy also to Sudan and Egypt.
The large investments in infrastructure undertaken in recent years have started bearing their fruits, as Ethiopia experienced a sustained rapid growth in the last decade, with real GDP growth averaging 10 percent annually. At the same time, poverty declined substantially and the provision of key public services has improved. The share of population with access to electricity, for instance, has increased from 14 percent in 2005 to 43 percent in 2016.
Significant gaps still remain, however. According to the Global Infrastructure Hub estimates, Ethiopia faces an investment shortfall of about USD 285 billion to achieve the targets set by the 2030 Agenda. But further investment has to be planned keeping in mind that the model used in the past to finance the infrastructure expansion is starting to show its limits in terms of debt sustainability, crowding out of private credit, and weak external competitiveness due to exchange rate appreciation (World Bank 2016). In particular, although the debt-to-GDP ratio declined from 107 percent in 2002 to 38 percent in 2009 (thanks to debt relief), since then it started increasing sharply and it reached 62 percent in June 2018. Even in presence of sustained economic growth, the adverse debt dynamics is reflected in the IMF and World Bank assessment, according to which Ethiopia is at high risk of debt distress (IMF 2018b). Moreover, absorptive capacity constraints could undermine the projects’ success rate and reduce the dividend of public investment (Presbitero 2018). As a result, investment has recently started declining and the large external imbalances and the public debt burden are constraining future growth.
The experience of Ethiopia, although unique in a number of respects, can be generalized to other developing countries, at least in regard to the trade-off between investment financing and debt sustainability. A key lesson for policy makers is that, even in presence of strong growth and large investment needs, any investment scaling up has to consider the risks that debt-financed public investment and higher public debt could pose on debt sustainability and future economic growth.
2. While there are good reasons to issue debt, there are also political failures that induce governments to borrow too much—leading, in some cases, to public debt levels that are hard to rationalize as the optimal decision of a benevolent social planner. Such excessive debt accumulation might not be without costs, as large debts could circumscribe future capacity to stabilize the business cycle—because of limited fiscal space—and possibly impair economic growth by either crowding out private investment or by increasing uncertainty about future tax and inflation rates. (Risks of public debt becoming unsustainable and leading to a debt crisis are discussed in another paper in Abbas, Pienkowski and Rogoff (2019)). This paper starts by discussing why governments borrow, separating good reasons for issuing debt (Section 2) from bad ones (Section 3) and it concludes by describing the link between public debt and economic growth (Section 4). Section 5 concludes.
II. Good Motives to Borrow
3. Budget deficits are the buffer that governments use to delink spending and revenues during certain periods of time. Why is it not always optimal to balance the budget? The principle of tax smoothing states that during periods of exceptionally high spending the government should run a deficit to finance those expenditures with future tax revenues. Periods of unusually high expenditure can be the result of various shocks (wars, natural disasters, etc.) or to spending decisions that reflect an expected economic, financial, or social benefit. Increasing spending during recessions can smooth the business cycle—raising employment, output, and incomes. Embarking on large public investment projects could lead to faster growth. Accommodating the need for structural reforms via short-run deficits can pay off in the future via higher GDP and higher tax revenues. In these cases, the principle of tax smoothing states that the government should run a deficit. At the same time, the government must be mindful that the expenditure will indeed reap the expected benefit—and that it is not undertaking the expenditure simply because there is easy financing available.
4. In this section we begin by elaborating on the principle of tax smoothing and then discuss some of the cases where attention needs to be paid to the returns on investment. We finish by presenting an additional argument for the government to issue debt—namely to provide the private sector with a safe asset.
A. The Logic of Tax-Smoothing
5. Governments have long financed extraordinary expenditures by issuing debt—most notably, when fighting wars.1 Often, it would have been socially and politically unacceptable to try to finance such a level of expenditure through contemporaneous taxation alone, so the government resorted to issuing debt. But there is also a sound economic rationale. If the government can only raise distortionary taxes, and if the cost of the economic distortion is convex (i.e., increasing at an increasing pace) in the tax rate, then it makes sense to try to “smooth” taxes over time in order to minimize the total distortionary cost.
6. One of the first to articulate the concept of tax smoothing explicitly was US Secretary of the Treasury, Albert Gallatin. In his annual Report to Congress, Gallatin (1807) argued:
It appears necessary to provide revenues at least equal to the annual expenses on a peace establishment, the interest on the existing debt, and the interest on loans that may be raised. [As to] whether taxes should be raised to a greater amount or loans be altogether relied upon for defraying the expenses of the war…the losses and privations caused by war should not be aggravated by taxes beyond what is strictly necessary. An addition to the debt is doubtless evil, but experience having now shown with what rapid progress the revenue of the Union increases in time of peace; with what facility the debt, formerly contracted, has been reduced; a hope may be confidently entertained that all the evils of war will be temporary and easily repaired; and that the return of peace will, without any effort, afford ample resources for reimbursing whatever may have been borrowed during the war.2
7. The idea was formalized by Barro (1979), who assumed a convex cost function, and showed that minimizing the present value of the distortionary burden involved equalizing the marginal cost of levying taxes over time. For a given tax base, this implies that the tax rate should be constant over time. If public expenditure is fixed, governments should run deficits and accumulate debt in bad times (when the tax base is cyclically low) and run surpluses and pay down debt in good times (when the tax base is high).3
8. While the logic of tax smoothing is clear, the problem that governments confront is that their expenditure is unlikely to be smooth. The classic example is that of a war: a (hopefully) temporary spike in expenditure far above the normal level of government spending. To finance such expenditure through current taxation may be prohibitively costly. It is even possible that as governments raise taxes to finance such a large spending, total tax revenues fall, making it physically impossible to finance the additional spending. This is what economists refer to as being on the wrong side of the Laffer curve.4
9. The argument is not just about large events like wars. In general, whenever there is some “lumpiness” in the government’s spending, there will be a divergence between the time path of expenditure and the (optimally) constant taxes—with deficits (and hence debt) making up the difference. In fact, it can be shown that optimally, the government’s overall balance should equal the present discounted value of expected future changes in government spending (Ghosh, 1995a). Thus tax smoothing applies only to temporary changes in spending: if there is a permanent increase or decrease in spending, then the expected change is zero, and the government should run neither a deficit nor a surplus. If the change in spending is permanent, then its burden cannot be spread over time to reduce the distortionary cost of the associated taxes. On the contrary, running a deficit in the face of a permanent rise in government spending would imply that—if the government is to respect its budget constraint (and not default)—then, eventually, it would have to raise taxes by even more to pay for both the higher spending and the interest on the accumulated debt thus violating the principle of tax smoothing.
10. Taking the logic one step further, if the government recognizes that it may face unexpected shocks (that raise spending above its normal level), it may want to “save for a rainy day” by accumulating assets or paying down debt (Aiyagari et al. 2002).5 Calibration exercises for the US government by Bhandari et al. (2016) suggest that in the long-run, the government should hold a positive, albeit small, net asset position as precautionary savings against future spending shocks. In general, the government will want to hold a portfolio of debt and financial assets that minimizes the risk that it will have to alter tax rates across time or states of nature (Bohn, 1990; Barro, 1995). Building on this premise, numerous academic papers have explored the optimal capital structure of the government’s assets and liabilities in a stochastic setting.6
11. Three further points bear emphasizing about the issuance of debt for tax smoothing purposes. First, for a given the level of output (GDP), unless the government is borrowing from foreigners, the issuance of public debt does not increase the resource envelope of the economy (the Keynesian models discussed below assume that there are occasions in which the deficit has a positive effect on GDP, at least in the short run, thus expanding the economy’s resource envelope). Thus, if the government issues domestic public debt in order to finance spending, then this debt necessarily crowds out private absorption (consumption or investment). Therefore, the only purpose of such borrowing would be to smooth taxes and thus lower the distortionary cost to the economy. But if the government borrows from foreigners—either directly or indirectly (i.e., issues debt to residents who in turn borrow from abroad)—then such borrowing would also expand the economy’s real resource constraint, allowing the private sector to also smooth consumption against shocks to government spending. In such cases there will be a positive association between the budget deficit and the current account deficit of the country—and, correspondingly, between the issuance of public debt and the issuance of (public or private) external debt.
12. Second, while the government’s ability to issue debt is welfare improving—inter alia, because it allows for tax smoothing—the debt itself becomes a dead-weight loss once issued. Even purely domestic debt—”a debt we owe ourselves”—represents an economic loss, equal to the present value of the economic distortions associated with the taxes necessary to repay it.7 (If the public debt is held externally, then there is an additional real resource transfer that will need to be made to non-residents.) The greater the inherited debt, the higher the taxes required to service it. If taxes are distortionary and fall on any factor (e.g., labor or private capital) that is complementary to the productivity of public capital, then—optimally—a government that inherits a higher level of public debt will undertake less public investment, with corresponding effects on output and growth (Ostry, Ghosh, and Espinoza, 2015).
13. Third, the logic of tax smoothing seems to apply regardless of the cost of borrowing (i.e., the interest rate the government must pay on its debt). Intuitively, however, for a given spending shock, a government will want to borrow less the higher the interest rate it must pay on its debt. In fact, this result follows from the tax-smoothing argument. The simplest example is to consider a situation in which output (the tax base) and government spending are constant. Taxes would then be set to equal government spending plus the interest on the stock of debt. In such a situation, if there was an unexpected, temporary increase in government spending (e.g., a war that lasts one year), then the government should run a deficit to help finance the higher spending. Taxes would (by tax-smoothing) immediately and permanently increase to the level required to pay for the normal level of spending plus the interest on the debt that will have been accumulated during the period of exceptional spending. Now, if the government faces a high interest rate on its borrowing, then taxes have to be raised correspondingly high. But since taxes are raised immediately to their permanent higher level, for a given spending shock, a government that faces a high interest rate on its debt would borrow less than a government that faces a low interest rate.
14. Summing up, the tax smoothing argument suggests that countries should accumulate public debt to finance large and lumpy expenditure (such as wars, natural disasters, and large investment projects), but also that debt accumulation during recessions should be accompanied by debt reduction in good times. While there is ample evidence that countries do accumulate debt during wars, tax smoothing is hard to reconcile with long-term debt accumulation during tranquil periods (the average debt-to-GDP ratio of the G7 countries went from about 40 percent in the 1970–1980 period to over 80 percent in 2007, see Figure 1). The link between debt accumulation and investment is even less clear. A simple regression that controls for country- and year-fixed effects shows a positive correlation between the debt-to-GDP ratio and public investment in advanced economies, implying that a one percentage point change in the debt-to-GDP ratio is associated with a 0.04 percentage point increase in public investment (Figure 2). This suggests that, typically, only a small percentage of debt issuance (4 percent) is used to finance public investment projects.8 It should be noted, however, that these are simple correlations, which do not allow for heterogeneity. Bacchiocchi, Borghi, and Missale (2011) find a negative correlation between debt and public investment in countries with a high debt ratio, and a positive correlation between debt and public investment in countries with low debt ratios. This may help explain the diversity of empirical results regarding the link between public debt and output growth, explored further below.
The Evolution of the Debt-to-GDP ratio in G7 countries
Citation: IMF Working Papers 2019, 101; 10.5089/9781498312103.001.A001
Source: Global Debt Dataset (Mbaye et al., 2018).
The Evolution of the Debt-to-GDP ratio in G7 countries
Citation: IMF Working Papers 2019, 101; 10.5089/9781498312103.001.A001
Source: Global Debt Dataset (Mbaye et al., 2018).The Evolution of the Debt-to-GDP ratio in G7 countries
Citation: IMF Working Papers 2019, 101; 10.5089/9781498312103.001.A001
Source: Global Debt Dataset (Mbaye et al., 2018).
Correlation Between Change in Public Debt and Contemporaneous Public Investment
Citation: IMF Working Papers 2019, 101; 10.5089/9781498312103.001.A001
Source: Global Debt Dataset (Mbaye et al., 2018) and World Economic Outlook.Notes: A regression of the ratio of public investment over GDP at time t against the change in the ratio of general government debt over GDP between t and t-1, controlling for year and country fixed effects, gives a coefficient on the debt variable of 0.041 (p-value of 0.011), meaning that a 10 percent increase of the debt-to-GDP ratio is associated with 0.4 percent lower ratio of public investment over GDP. To generate the binned scatterplot, starting from the sample of 19 OECD economies (data on general government for New Zealand are not available), the change in the ratio of general government debt over GDP between year t and t-1 (x-axis) and public investment (as a percent of GDP, y-axis)) in year t are regressed against year and country fixed effects. Then, the x-residuals are grouped into 50 equal-sized bins and the chart plots, for each bin, the mean of public investment (as a percent of GDP) in year t, within each bin, holding the controls constant. The red line is the linear fit of the OLS regression of the y-residuals on the x-residuals. The number of observations is 899.
Correlation Between Change in Public Debt and Contemporaneous Public Investment
Citation: IMF Working Papers 2019, 101; 10.5089/9781498312103.001.A001
Source: Global Debt Dataset (Mbaye et al., 2018) and World Economic Outlook.Notes: A regression of the ratio of public investment over GDP at time t against the change in the ratio of general government debt over GDP between t and t-1, controlling for year and country fixed effects, gives a coefficient on the debt variable of 0.041 (p-value of 0.011), meaning that a 10 percent increase of the debt-to-GDP ratio is associated with 0.4 percent lower ratio of public investment over GDP. To generate the binned scatterplot, starting from the sample of 19 OECD economies (data on general government for New Zealand are not available), the change in the ratio of general government debt over GDP between year t and t-1 (x-axis) and public investment (as a percent of GDP, y-axis)) in year t are regressed against year and country fixed effects. Then, the x-residuals are grouped into 50 equal-sized bins and the chart plots, for each bin, the mean of public investment (as a percent of GDP) in year t, within each bin, holding the controls constant. The red line is the linear fit of the OLS regression of the y-residuals on the x-residuals. The number of observations is 899.Correlation Between Change in Public Debt and Contemporaneous Public Investment
Citation: IMF Working Papers 2019, 101; 10.5089/9781498312103.001.A001
Source: Global Debt Dataset (Mbaye et al., 2018) and World Economic Outlook.Notes: A regression of the ratio of public investment over GDP at time t against the change in the ratio of general government debt over GDP between t and t-1, controlling for year and country fixed effects, gives a coefficient on the debt variable of 0.041 (p-value of 0.011), meaning that a 10 percent increase of the debt-to-GDP ratio is associated with 0.4 percent lower ratio of public investment over GDP. To generate the binned scatterplot, starting from the sample of 19 OECD economies (data on general government for New Zealand are not available), the change in the ratio of general government debt over GDP between year t and t-1 (x-axis) and public investment (as a percent of GDP, y-axis)) in year t are regressed against year and country fixed effects. Then, the x-residuals are grouped into 50 equal-sized bins and the chart plots, for each bin, the mean of public investment (as a percent of GDP) in year t, within each bin, holding the controls constant. The red line is the linear fit of the OLS regression of the y-residuals on the x-residuals. The number of observations is 899.B. Keynesian Demand Stimulus
15. One example of government spending increasing during a transitory period is the type of countercyclical fiscal policy many governments implement during recessions. While in many ways this fits our previous logic of tax smoothing, it adds a second dimension and justification for that spending. The discussion thus far has taken output as given and considered optimal fiscal policy for an exogenous path of GDP. But during recessions, governments also try to influence output, so a full assessment of the benefits of running deficits must take account of this aspect as well.
Why Countercyclical Policy is Needed
16. In most macroeconomic models, monetary and fiscal policy are effective tools to stabilize the business cycle. In open economies, the Mundell-Fleming results imply monetary policy will be ineffective under fixed exchange rates and an open capital account, so only fiscal policy is available. The traditional Keynesian IS-LM model provides the basic intuition behind the standard prescription for countercyclical policies in order to stabilize output. In the case of fiscal policy, changes in spending and taxes help stabilize aggregate demand by acting as a counteracting force to changes in private spending. More sophisticated (New Keynesian) models can also validate the IS-LM intuition in dynamic and optimizing environments (Beetsma and Jensen, 2005). This intuition is the basis of most policy discussions on the need for countercyclical policy (International Monetary Fund, 2008).9
17. In discussing output stabilization, there is a sense in which fiscal policy and monetary policy can be seen as substitutes in these models. If a choice is to be made, then monetary policy has traditionally been seen as quicker and less subject to political interference relative to fiscal policy. But this logic does not quite apply to endogenous changes in fiscal policy (what is known as automatic stabilizers), which is why they have typically been seen as superior to discretionary policy (Taylor, 2000).
18. But even if monetary policy is faster and potentially more effective than fiscal policy in stabilizing fluctuations, there may be instances when monetary policy cannot achieve the first- best result, even with a flexible exchange rate.10 In particular, when monetary policy is constrained by the zero lower bound on interest rates—as was the case for many central banks during the Global Financial Crisis—the burden necessarily falls on fiscal policy to be the stabilizing tool (Eggertsson and Woodford, 2004). More generally, in the presence of more than one distortion in the economy, not just price rigidity, monetary policy alone may not suffice to bring the economy to its first-best outcome, and fiscal policy could play a role in allowing policy makers to reach first best (Blanchard and Gali, 2010).
19. The normative statement that fiscal policy should be countercyclical is not always borne out empirically, however. Evidence of procyclical fiscal policy is stronger among emerging markets and, in particular, among Latin American economies, as documented by Gavin and Perotti (1997) and by Kaminsky, Reinhart, and Végh (2004). In part, this may be because the scope for deficit financing during downturns is more limited in such countries, while the lack of fiscal discipline during the upswing may reflect political economy considerations (as discussed in Section 3, below). It is also the case, however, that recent evidence—particularly in the aftermath of the Global Financial Crisis—is more encouraging (Frankel, Vegh, and Vuletin, 2013).
20. The evidence for OECD or European economies is somewhat mixed. Most times fiscal policy is countercyclical but occasionally it turns procyclical, as in the case of recent fiscal consolidations in European countries (Égert, 2012; Fatás, 2018). One problem is judging the output gap in real time; especially after a financial crisis, the trajectory for potential output—and hence the output gap—may have changed. Procyclical policy has negative economic consequences as it leads to higher output volatility and lower growth (Aghion et al. 2007).
Countercyclical Policy, Deficits and Debt
21. How do we characterize countercyclical fiscal policy? We judge fiscal policy by its potential effects on output growth, what is typically referred to as a measure of the fiscal policy stance. Measuring these effects requires the use of an economic model, and different models might lead to different conclusions. In a classic paper, Blanchard (1993) discusses this issue at length. By studying a variety of models, he concludes that a good indicator of the fiscal policy stance is the change in the inflation-adjusted budget balance as a ratio to GDP. The logic is that spending affects aggregate demand while taxes help stabilize disposable income and therefore private spending. The effects of taxes and spending might not be identical, but the budget balance comes close enough to capture their combined effect.
22. Blanchard (1993) logic is commonly used in policy discussions where changes in the budget balance are used as an indicator of the fiscal policy stance. This establishes a direct connection between countercyclical fiscal policy and debt. When growth is below trend, governments will run deficits, and thus debt will accumulate.
23. When looking at the change in the budget balance it is important to distinguish between automatic and discretionary changes, even if from the perspective of aggregate demand this is largely irrelevant—it is the overall balance that matters. Automatic stabilizers capture changes in the budget balance that are the result of tax or spending laws that were not decided or modified as a result of current economic conditions. What types of tax and spending rules generate stronger automatic stabilizers? Given our logic on the fiscal policy stance, the answer is those that generate a larger swing in the budget balance. For this, we do not necessarily require strong cyclicality in taxes or spending. In fact, the largest source of automatic stabilizers in advanced economies is acyclicality of public spending. If the government maintains spending constant when GDP is falling, then even if taxes are proportional (so there is no extra countercyclicality in the tax schedule), deficits will increase. In this stylized case, the magnitude of the automatic stabilizers is simply proportional to size of government. Larger governments will have stronger stabilizers and run larger deficits during downturns. The data show that the majority of automatic stabilizers among advanced economies comes from this effect (Fatás and Mihov, 2012). Of course, in addition, spending might automatically increase during downturns and crises, and certain taxes have an elasticity larger than one, both of which add to the strength of automatic stabilizers.
24. Beyond automatic stabilizers, governments also engage in discretionary fiscal policy changes. These changes follow the same logic as they also contribute to deficits and accumulation of debt during downturns. There is evidence, among advanced economies, that discretionary fiscal policy is used more aggressively in those countries that have the weakest automatic stabilizers (because of their smaller government size)—highlighting the substitutability between these two types of fiscal policy (Fatás, 2009).
From Cyclical Deficits to Accumulation of Debt
25. Our argument thus far is that fiscal policy should be used as a stabilization tool. If governments plan properly for the right balance over the business cycle, then this behavior should have no effect on government debt over the long run.11 Debt may be expected to increase during periods of below-average economic growth and then decrease when growth is above-average. However, as discussed in the previous subsection, public debt levels were increasing in many advanced economies—despite the booming world economy—even before the onset of the Global Financial Crisis and the ensuing Great Recession.12 Can this trend be at all related to the dynamics of stabilization policy over the cycle? Is there an asymmetry? And if there is, why are governments not able to produce policies that are more symmetric?
26. We can think of two potential hypotheses that lead to asymmetries. First, while governments are ready to apply countercyclical policies during recession, they are less likely to follow the same logic during expansions. This is related to our earlier argument about observed procyclical policies (or not enough countercyclical policies). Empirically this is the case, at least for some countries (see Fatás and Mihov (2010) for a sample of European countries or Alesina, Campante, and Tabellini (2008) for a larger sample). The next section discusses the political distortions that may lead to such asymmetry.
27. The second argument is not so much about political incentives of governments but about excessive optimism or pessimism when forecasting GDP growth. For example, during periods of strong growth, governments produce forecasts of potential output growth that are too optimistic. Fiscal policy is planned under the assumption of no significant economic downturn resulting in fiscal policy being too procyclical in good times. The data confirms this hypothesis as estimates of potential output and its growth rate tend to be highly procyclical and this leads to excessive expansionary fiscal policy in good times (Mc Morrow, Roeger, and Vandermeulen, 2017). As an example, in December 28, 2000, President Clinton announced that the US was on course to eliminate its government debt within the following 10 years. The macroeconomic scenario supporting this forecast did not include the 2001 and 2008 recessions that happened within the next 10 years (nor, in fairness, did it foresee the Afghanistan and Iraq Wars).13
28. But is this bias in growth forecasts only present in good times? Not quite; we also observe excessive pessimism during downturns when potential output estimates are revised downwards. This generates a bias towards excessively tight fiscal policy during recessions. In principle, this bias is towards lower-than-optimal deficits and debt, potentially counteracting the excessive optimism during expansions. Given the difficulties in predicting turning points, however, lags in the implementation of discretionary fiscal measures could result in a net bias toward larger deficits and higher debt on average.
29. But if excessive optimism in good times is matched by excessive pessimism in bad times we are back at a symmetric argument and we should not observe debt trending upwards or downwards. There are however two possible additional asymmetries that can generate a drift of debt ratios:
There could be an interaction between procyclical GDP forecasts and the political economy argument discussed in the next section. The procyclical forecast bias during expansions might be acted upon while the procyclical forecast bias during recessions is ignored. During recessions the lack of political incentives to engage in fiscal consolidations might not be there and pessimistic growth scenarios do not lead to lower deficits. If this is the case, the overall bias will be towards higher debt.
But there is also an asymmetry in terms of the way the economy reacts to procyclical fiscal policy as fiscal policy multipliers tend to be larger during recessions than booms (Freedman et al., 2010; Auerbach and Gorodnichenko, 2013; Jordà and Taylor, 2016). As a result, the procyclical nature of fiscal policy will cause more damage to GDP during downturns than during expansions. This is especially the case during deep recessions where monetary policy may be constrained by the zero-lower bound (and unable to stabilize output). In these cases, excessive pessimism about potential output and the associated fiscal policy contraction will make the recession much deeper. And the worst-case scenario is when there is hysteresis (i.e., permanent effects of cyclical shocks). Now the negative effects on GDP are likely to become permanent, validating the unfounded pessimistic expectations of governments. These dynamics can lead to a result that runs diametrically opposite to the government’s objectives. Under some scenarios governments that engage in contractionary fiscal policy to reduce debt ratios might end up instead with higher debt-to-GDP ratios. This is what the literature calls self-defeating fiscal consolidations (see Fatás, 2018; or DeLong and Summers, 2012). In this case, the resulting bias is again towards more debt even if governments were too conservative from a fiscal policy point of view because of their pessimistic views on GDP growth. Unlike in the previous cases, however, the solution here is for a more aggressive policy (larger deficits during crises) to avoid the negative effects on GDP.14
30. Finally, it bears emphasizing that accumulation of debt in times of crisis is not solely the result of standard Keynesian countercyclical policy but also of the support that governments may need to provide to repair weak financial and banking systems. In many cases this support results in changes in government debt levels that are as large, or larger than, the effects of demand-supporting fiscal policy measures (International Monetary Fund, 2015, Campos, Jaimovich, and Panizza, 2006). This is important because, even if potential GDP forecasts are unbiased, estimates of deficits and debt will be too optimistic if they do not consider the occasional support for the financial system during large crisis. Once these—hopefully rare—events happen, debt levels will be higher than expected. At that point, the logic of tax smoothing that we have discussed earlier implies that debt levels will remain higher for a long time as the adjustment is optimally spread over many years (Ostry, Ghosh, and Espinoza, 2015).15
C. Long-Term Investment and Deficits
31. Investment projects financed by the government fall into the category of lumpy spending that, according to our tax-smoothing argument, should be financed by deficits. But unlike other forms of spending (i.e., government consumption), here the government is acquiring an asset that will deliver services in the future. This strengthens the argument for financing investment with debt.16
32. The welfare effects of debt-financed investment will depend on the social returns of the projects. Although this is also true for any other form of government spending, investment is typically assumed to deliver a social or even a financial return that justifies the spending, which might even generate the taxes necessary to repay the debt in the future. The greater the reliance on some nebulous social benefit in making the case for the investment, the greater the risk that the project turns out to be a white elephant. More generally, limited absorptive capacity could constraint the growth dividend of additional public investment projects, especially in periods of rapid acceleration of public investment (Presbitero 2018)— increasing the likelihood that the government later runs into debt-servicing difficulties (see Box 1).
33. There are other instances where the government might also be acquiring assets and issuing debt to finance its purchases. Financial sector bailouts might lead to an increase in the balance sheet of governments (Reinhart and Rogoff, 2009; Laeven and Valencia 2013; Amaglobeli et al., 2017). When governments need to recapitalize the banking system, they acquire a financial asset (the equity stake in the bank), which they typically finance by issuing debt (in fact, the recapitalization often takes the form of a government bond). In this case we potentially have several arguments that justify these decisions. Stabilizing the financial sector can lead to a reduction in the severity of the decline of GDP. Not only will this be welfare enhancing, it will also raise tax revenues for the government. In addition, the assets acquired might be undervalued by the market because of panic. In this case the assets could deliver a return that will partly or fully compensate the financial costs of issuing debt.
34. A recent study on the fiscal costs of systemic banking crises over the period 1970– 2011 shows that the median cost of direct government intervention in the banking sector amounted to about 7 percent of GDP (factoring in the indirect fiscal costs raises the impact of banking crises to 12 percent of GDP, International Monetary Fund, 2015).
35. A final case where deficits can be justified even if the government is not acquiring a physical or a financial asset is to support structural reforms. The political economy of structural reforms means that it is very difficult to find support for them in democratic environments (Fernandez and Rodrik, 1991). This difficulty is the result of the uncertainty about who are the winners and the losers of those reforms. One way to ensure support would be through spending or tax measures that make those benefits immediate and reduce the potential uncertainty about the long-term benefits. In this case the government is investing in the necessary credibility needed to deliver structural reforms that will produce a social and even financial return in the long run (Banerji et al., 2017).
D. Asset Management and Government Debt as Safe Asset
36. In the previous section we emphasized the importance of looking at the asset side of the government’s balance sheet to understand the existence of debt. A particular case that has become more relevant in recent times is the need for governments to provide a special asset. For example, many governments have external debt but also hold foreign exchange reserves because they provide FX liquidity at normal times or at a time when it is most needed (e.g., during a sudden stop of capital flows or an export shortfall or a terms-of-trade shock).
37. A second need for a special asset comes from financial markets and the need for a safe asset. Public debt in this case is not issued to meet the government’s borrowing needs but to provide financial markets with risk-free instruments. Historically (e.g., development of Britain’s financial markets, including the stock exchange, money markets etc.), the government debt market has been important for financial market development.17 Indeed, at the national level, government debt markets have often played a key role in developing nascent financial markets, including extending the yield curve to longer maturities and providing a benchmark. International Monetary Fund (2012) highlights the overall benefits of safe assets. Abbas and Christensen (2010) show that moderate levels of non-inflationary government debt have a positive overall impact on economic growth. Gorton and Ordoñez (2013) analyze the benefits of government debt as a safe asset during crisis to show that within their model “The decline in output during a crisis is lower to the extent that there are more government bonds outstanding”.18
38. At the international level, there is a similar need for safe assets, and these assets are likely to be associated to one of the major reserve currencies (US dollar, Euro, Yen). In fact, the Global Financial Crisis—because of a combination of flight-to-safety and several sovereigns losing their AAA status—has resulted in a shortage of global safe assets with a variety of consequences (see Caballero, Farhi, and Gourinchas, 2008; Brunnermeier et al., 2017; Gourinchas and Jeanne, 2012, among others).
E. Dynamic Inefficiency
39. A final potential argument for the issuance of government debt is the possibility that the economy is dynamically inefficient and that the private sector cannot optimally provide vehicles to transfer wealth across generations. Government debt can play that role. In this environment, issuing additional government debt not only can be sustainable but it is optimal (Blanchard, 1985, 2019). For dynamic inefficiency to hold, it requires that the rate of return of an economy must be below its growth rate. Interest rates on government debt are often below GDP growth rates, but what matters is the rate of return on capital.19 In a seminal study, Abel et al. (1989) provided strong evidence for 6 advanced economies that the criterion for dynamic inefficiency was not met. However, recent decades have seen substantial reductions in real interest rates on safe assets, which could suggest it could be worth revisiting their findings to see whether their conclusions still hold. Geerolf (2017), concludes that dynamic inefficiency cannot be ruled out for several advanced economies. But whether the evidence is sufficiently compelling to warrant a clear policy recommendation in some of these countries remains an open question (Blanchard and Summers, 2017; Blanchard, 2019).
III. Bad Reasons to Issue Debt
40. The previous section showed that there are good reasons to borrow. Budget deficits, and the resulting accumulation of debt, can be optimal during recessions or in the presence of exceptional events such as war, natural disasters, or financial crises. Borrowing may also be justified by the need to finance large investment projects.
41. However, cyclical stabilization should not lead to a steady accumulation of debt, as budget deficits during bad times should be offset by surpluses during economic expansions. Moreover, while countries do accumulate large debts during wars, there is only limited evidence of a link between public debt accumulation and surges of public investment (Figure 2).
42. A benevolent social planner would borrow up to the point at which the social marginal cost of an additional unit of debt (this includes principal, interest repayment, and any possible externality brought about by higher debt levels) equals the social return of an additional unit of debt-financed government expenditure. Overborrowing refers to a situation in which the government borrows more than is socially optimal. Yared (2018) suggests that the accumulation of public debt in recent decades is due to overborrowing driven by political distortions which leads to time inconsistent preferences and a bias towards present consumption.
A. Why Do Countries Overborrow?
43. Just as it is wrong to compare the behavior of the government with that of a household— because the government is a large player and its borrowing decisions can have important spillovers, positive and negative, on the economy—it is also wrong to assume that policymakers always try to maximize social welfare. While it is reasonable to assume that the decision of a household head to borrow aims at maximizing the household’s welfare, a policymaker’s decision to contract public debt may be less benign. In order to understand why countries overborrow, it is necessary to move from normative to positive theories of public debt (Alesina and Tabellini, 1992). 20
44. The economics literature has emphasized four potential sources of excessive debt accumulation: (i) political budget cycles and rent seeking; (ii) intergenerational transfers; (iii) strategic manipulation; and (iv) common pool problems. It should be highlighted that while many politicians and policymakers are well-intentioned and attempt to maximize social welfare, the economics literature tends to model politicians and policymakers as self-interested agents.
45. We start with a flash review of the literature on the political economy of debt, we then discuss how budget institutions can mitigate political failures that lead to excessive debt accumulation, and conclude with a discussion of sources of debt that are not directly related to standard fiscal policy.21
Political Budget Cycles and Rent Seeking
46. The literature on the political budget cycle suggests that politicians cut taxes and increase spending to increase the likelihood of being reelected. At the most basic level, the presence of a political budget cycles requires that voters suffer from “fiscal illusion.” Only voters who do not fully understand the intertemporal nature of fiscal policy may be tempted to vote for politicians who cut taxes or provide more public services without increasing taxes. In the traditional public choice literature, fiscal illusion is amplified by the asymmetric application of Keynesian stabilization policies, with policymakers happy to run budget deficits during recessions but less inclined to run surpluses in period of rapid economic growth (Buchanan and Wagner, 1977).
47. Political budget cycles models do not necessarily require the presence of irrational voters. For instance, Rogoff and Sibert (1988) develop a model in which the presence of fully rational but imperfectly informed individuals leads to a political business cycle because pre-electoral budget deficits are the only mechanism through which policymakers can signal their competence (the reason being that only competent politicians will be able to balance the budget after the election). Another implication of these types of models is that policymakers may engage in more visible, but not necessarily more efficient, types of public expenditure (Rogoff, 1990).22
48. Political business cycle models implicitly assume that policymakers want to remain in power. This may be because of ego-rents or because, by staying in power, they can implement their favorite policies. It is, however, also possible that policymakers want to be in power to extract resources from the economy. Yared (2010) studies the behavior of politicians that try to extract rents and need to decide whether to extract a limited amount of resources in each period or extract everything they can in one period and then lose power. One of the implications of the model is that a high level of debt reduces the politicians’ incentive to extract the maximum amount of rent and makes her behave more likely a social planner.
Intergenerational Transfers
49. Individuals can leave positive bequests to their offspring, but private negative bequests cannot be enforced by law. However, individuals who would like to leave a negative bequest can use public debt to redistribute resources from future to current generations. Cukierman and Meltzer (1989) use an overlapping generations model to study an economy with two types of individuals: citizens who would like to leave a positive bequest to their children and constrained citizens who would like to leave a negative bequest. The first group only cares about public debt through its effect on the economy because individuals who belong to this group can fully undo the intergenerational effects of higher public debt by increasing their bequest. Bequest-constrained individuals, instead, would like to issue more debt because this relaxes their constraint. In such a set-up, the level of debt depends on the bequest constraint faced by the median voter.23
50. Other models that focus on intergenerational transfers include Tabellini (1991) who develops a model with defaultable debt and wealthy and poor voters. In this setting, higher levels of debt (up to a point) create an incentive to repay the debt by linking intergenerational with intragenerational transfers. Song et al. (2012), instead, study a model in which the young and the old have different preferences for public goods and taxation is distortionary. In this this setting, the level of debt is determined by these preferences (that can vary across countries) and the political power of the two groups. Jackson and Yariv (2015) show that if there are two groups of individuals and one group (the old) cares less about the future than the other group (the young) a government that aggregates the preferences of these two groups may suffer from a present bias. Yared (2018) shows that theory is consistent with the fact that there is a positive cross-country correlation between the growth rate of public debt and aging of the population. It is worth noting that standard social planner model would predict the opposite correlation.
Strategic Manipulation
51. On February 18, 1981 President Reagan described his program for economic recovery in a joint session of the US Congress. Among the topics discussed in his speech, there was the high level of public debt, which was approaching $1 trillion (this was total US Federal debt, Federal debt held by the public was about $770 million or 25 percent of GDP).24 Eight years later, the US federal debt held by the public had surpassed $2.1 trillion (a 100 percent increase in real terms) and reached 39 percent of GDP.
52. Why would a conservative like president Reagan accumulate so much debt, and why are large and persistent primary surpluses often associated with left-of-the center governments (Eichengreen and Panizza, 2016)? Persson and Svensson (1989) show that, in the presence of two parties with different preferences for spending and taxation, left-of-the-center parties (which prefer more public goods at the cost of higher taxes) may decide to run budget surpluses so that the right-wing party will inherit a low level of debt and will not have a strong incentive to reduce public expenditure. Similarly, the right-wing party will increase the level of debt so that the left wing party will have to limit spending when in power.25
53. While the model of Persson and Svensson (1989) shows how debt can be used to influence the actions of successive governments, it does not necessarily lead to excessive debt accumulation because deficits by right wing governments are canceled with surpluses by left wing governments. Alesina and Tabellini (1990) develop a model in which political parties have preferences for different types of public expenditure and accumulate debt in order to constrain the choices of future governments. In this setting, the level of debt depends on the likelihood of being reelected. Governments which are sure to stay in power behave like a social planner and issue no debt. However, governments with low probability of reappointment will overborrow.
54. The key intuition of the models of Persson and Svensson (1989) and Alesina and Tabellini (1990) is that debt is a state variable that the party in power can use strategically to influence and constrain the actions of successor governments.26 One weakness of these models is that they were developed in a set up that rules out output shocks and hence the tax smoothing motive that create incentives to self-insure by accumulating assets (Aiyagari et al., 2002). It is thus difficult to use these models to build testable predictions on how debt reacts to economic shocks in different institutional environment.
Common Pool
55. Common pool problems originate from the presence of externalities which lead to a situation in which the private benefit of an additional unit of public expenditure is different from the social marginal cost of funding this extra unit (Olson, 1965, and Ostrom, 1990). The presence of concentrated interests amplifies the common pool problem. When policy actions benefit a certain group and are funded with a general tax, the relatively small group of people who benefit from the policy will have strong incentives to lobby in favor of the policy (for a classic analysis of the role of pressure groups see Buchanan and Tullock, 1962). The much larger, but dispersed, group of actors that bears the cost of this action will have weaker incentives to act against it.
56. Common pool, problems may explain why fiscal adjustments based on spending cuts, albeit desirable are often hard to implement. For instance, Mauro and Villafuerte (2013) show that, while almost 90 percent of fiscal adjustment plans implemented by EU countries envisaged large spending cuts partly compensated by lower taxes, the data reveal that expenditure cuts often did not materialize and that this lack of expenditure cuts either resulted in smaller budget surpluses (or outright deficits) or was compensated by temporary measures aimed at increasing revenues. This is clearly suboptimal as there is ample evidence that expenditure-based fiscal adjustments are preferable and more likely to be long-lasting than revenue-based fiscal adjustments.
57. The exact way in which the common pool problem manifests itself depends on the institutional setting. There is a large literature in in political sciences (dating back to Weingast, Shepsle, and Johnse, 1981 and Baron and Ferejohn, 1989) that models common pool problem and pork barrel spending in the US Congress. However, common pool problems also apply to situations in which the budget law is prepared by the government and then sent to the legislative body for approval. In such a setting, it is possible to think of a strategic interaction between the Ministry of Finance, which worries about the overall budget constraint, and the spending ministries which are subject to pressure from different interest groups (Alesina and Perotti, 1996). In such a setting, hierarchical rules in which the Ministry of Finance first decides the overall budget envelope and then the line ministries decide on the allocation may play a role in reducing excessive spending (more on this below).
58. Common pool problems, however, may lead to overspending but not necessarily to budget deficits and debt accumulation, as the increase in spending could be matched by higher taxes. However, if property rights are not well defined and each group fears that any residual government asset will be appropriated by the other group, each group may find optimal to demand large transfers and push the government to its borrowing limit (Tornell and Lane, 1999 and Velasco, 2000).27
59. Political turnover amplifies common pool problems because if parties have different preferences for different types of public goods they will have an incentive to overspend in their favorite good when in power and this incentive to overspend is inversely related to the probability that the party will be in power in the next period. In other words, government that are not sure whether they will be in power next period are more likely to be impatient (Aguiar and Amador, 2011)
60. The empirical evidence is generally consistent with common pool models as it finds that budget deficits tend to be larger in countries characterized by deeper political cleavages and party fractionalization (Woo, 2003).
61. Common pool problems can also lead to overborrowing if legislators do not know whether they will be part of future governing coalitions. Battaglini and Coate (2008) show that adding uncertainty to a dynamic common pool model leads to two contrasting forces which unify the main findings of the normative literature on public debt (e.g., Barro, 1979 and Aiyagari et al., 2002) with those of the positive literature that emphasizes the role of political failures (e.g., Alesina and Tabellini, 1990). One the one hand, there is a self-insurance incentive: policymakers want to accumulate assets in order to insure against future shocks (as in Aiyagari et al., 2002). On the other hand, there is a political distortion: policymakers accumulate debt because they may not be part of future governing coalitions and higher levels of debt constrain the behavior of future policymakers as in the strategic models described above. When debt levels are low, political distortions dominate the self-insurance incentive and the government overborrows. As debt increases, the self-insurance motive becomes more important and fiscal policy becomes similar to the policy that would be chosen by a social planner (albeit with a higher equilibrium level of debt).28
B. Controlling Overborrowing
62. The economics literature identifies three possible avenues to limit overborrowing. The first focuses on the electoral system, the second on fiscal rules, and the third on budgetary institutions.
Electoral Systems
63. Battaglini (2010) shows that a simplified version of the model of dynamic electoral competition discussed in Battaglini (2014) yields the unambiguous prediction that proportional electoral systems suffer from a deficit bias with respect to majoritarian electoral systems. This prediction is in line with a large number of papers that show that democracies with a proportional electoral system accumulate more debt than democracies with a majoritarian system (e.g., Roubini and Sachs, 1989 and Grilli et al., 1991).29
64. A related literature that compares presidential and parliamentary democracies finds that presidential democracies tend to have smaller governments than parliamentary democracies (and, within parliamentary democracies, majoritarian systems have smaller governments than proportional systems) and that in parliamentary democracies increases in government spending during recessions are less likely to be reversed during economic expansions (Persson and Tabellini, 2003, 2004). If taxes remain constant over the business cycle, this behavior may lead to a ratchet effect and to a deficit bias in parliamentary democracies.
Fiscal Rules
65. Fiscal rules aim at addressing the time inconsistency problem and limit debt accumulation by imposing an upper limit on budget deficits. A government that implements a fiscal rule trades off a constraint on its own action (something the government does not like) with a constraint on successor governments (something the government does like). Fiscal rules have become increasingly popular: while in the mid-1990s there were fewer than 20 countries with a national or international fiscal rule there are now nearly 100 countries that adhere to some type of fiscal rule.30
66. The most extreme fiscal rule is the balanced-budget rule requiring zero deficits in every period. Such a rule may reduce welfare because it limits the government’s ability to use countercyclical policies (or to smooth taxes).31 A rule that aims at balancing the budget over the business cycle addresses this issue at the cost of being less transparent. Yared (2018) presents a detailed survey of these tradeoffs by discussing the role of public information, the degree of enforcement (including the role of escape clauses), and the costs and benefits of rules based on specific targets (i.e., the total or primary deficits) vis-à-vis rules that concentrate of policy instruments (such as spending).
67. On the empirical side, there is a large literature on the effect of balanced-budget rules for subnational governments (especially US states see, for instance, Poterba, 1994) and also a large literature on the fiscal rules adopted by many European countries. The results of this latter literature are mixed. On the one hand, Debrun et al. (2008) and Bergman et al. (2016) find that fiscal rules play a significant role in limiting budget deficits in European countries; on the other hand, Von Hagen (2006) suggests that the fiscal rules imposed by the Maastricht Treaty did not constrain the behavior of the largest countries in the euro area. The main challenge is to go beyond simple correlations and establish whether such rules have a causal effect on fiscal outcomes (Heinemann et al. (2018)). Caselli and Wingender (2018), using an innovative identification strategy and a bunching estimation method, find that the Growth and Stability Pact has led to a bunching of fiscal deficits around the 3 percent Maastricht deficit ceiling.
Budgetary Institutions
68. The preparation of the budget is a complex exercise that involves several players within the government as well as the interaction between the executive and the legislative. There is evidence that the institutions that regulate the preparation of the budget and guarantee its transparency have an impact on fiscal outcomes.
69. When focusing on the preparation of the budget, the economic and political science literature differentiates between hierarchical and collegial rules. The former tend to give more power to the ministry of finance, while the latter are more inclusive and give more power to the spending ministries and allow the legislature to amend the budget. Hierarchical rules mitigate the common pool problem and are thus associated with smaller deficits and debt accumulation (for surveys of the literature see Eichengreen et al., 2011, and Hallerberg et al., 2009). There is, of course, a trade-off in terms of democratic accountability.
70. Transparency of the budget also matters. Rogoff and Sibert (1988) emphasize that imperfect information can lead to political business cycles and Milesi-Ferretti (1997) discusses how politicians who want to overborrow have incentives to window-dress their budget laws, even more so when the politicians are corrupt.32 Standard strategies for manipulating the budget include keeping various items off-budget and adopting overoptimistic projections on either the state of the economy or on the effect of certain policies on tax revenues or expenditure.
71. Building on the intuition of Alesina and Tabellini (1990) and Rogoff and Sibert (1988), Beetsma et al. (2017) develop a model which finds that transparent budgets mitigate incentives to overborrow. This prediction is consistent with the empirical literature that finds that fiscal transparency is associated with lower levels of public debt in advanced, emerging market, and low-income countries (Alt and Lassen, 2006, Alesina et al., 1999, Dabla-Norris et al., 2010).
C. The Unexplained Part of Public Debt
72. So far we assumed that fiscal policy is the only driver of debt accumulation and that the stock of debt is equal to the sum of past budget deficits. However, debt accumulation can be better described as the sum of past deficits plus an unexplained residual, sometimes referred to as the stock-flow reconciliation (or adjustment).33 In some cases, this residual can be very large.
73. Consider the case of Uruguay, at the end of 2001, its net debt-to-GDP ratio was 35 percent (gross debt was 55 percent of GDP) and at the end of 2002 it had reached 76 percent (gross debt 106 percent). Yet over 2002 Uruguay’s total budget deficit was only 3.7 percent of GDP, the growth in debt was 17 percentage points of GDP higher than the deficit. The case of Argentina is even more striking. At the end of 2001, Argentina’s gross public debt was 49 percent of GDP, by the end of 2002 it had reached 152 percent of GDP while the 2002 public deficit was recorded at just above 2 percent. The debt grew 101 percent percentage points of GDP more than the deficit.
74. Campos et al. (2006) show that sudden debt explosions which cannot be fully explained by recorded deficits are a not limited to the few cases described above, and Cafiso (2012) shows this phenomenon is not limited to emerging market countries and that that the stock-flow reconciliation accounted for nearly one third of public debt growth in EU countries over 2008–10.34
75. The main drivers of this “unexplained” part of debt are balance sheet effects linked to the presence of foreign currency debt (Eichengreen et al., 2005, currency depreciations in the presence of dollar debt explain the debt explosions of Argentina and Uruguay mentioned above), banking crises (Amaglobeli et al., 2017), hidden deficits (and thus borrowing) driven by governments’ incentives to misreport public expenditure (Weber, 2012, shows that more transparent budgets are correlated with lower stock flow adjustments), and all sort of contingent liabilities linked to implicit subsidies and public guarantees.
76. While some of these fiscal risks are unavoidable, policymakers could implement policies aimed at mitigating them. For instance, a safer (from the point of view of the borrower) debt structure could mitigate some of the risks linked to balance sheet effects, and contingent debt instrument, such as GDP indexed bonds, could reduce fiscal costs at time of crisis (Borensztein and Mauro, 2004). The question is why these safer debt instruments tend to be underused.
77. Part of the answer has to do with the fact that local currency or indexed debt tends to be more expensive than what a fair insurance would predict (this is both because of market failures and because of the incentives linked to local currency debt as highlighted by Calvo, 1988, and Tirole, 2003). However, political failures also play a role as domestic currency debt and contingent instruments more generally work as an insurance policy: their cost must be paid upfront but their payoff may not occur until years later (or never occur). As such they tend to be poorly understood by the public and may not be desirable for policymakers who may be afraid to pay the cost of an instrument that might end up benefitting their successors (for a detailed discussion see Borensztein et al, 2006).
IV. Debt, Growth and Investment
78. Regardless of the motives to borrow, high levels of government debt can have adverse effects on the economy, as they may limit the capacity to run counter-cyclical fiscal policy and reduce private sector investment through the standard crowding-out channel; by tightening credit constraints; by creating the expectation of higher future distortionary taxation; or by increasing uncertainty. On the other hand, public borrowing—even if it results in a higher debt ratio—can be good for growth if the additional borrowing is directed at financing productive investment—such as a big push in infrastructure—or at stimulating aggregate demand (see above, Section 2.3), although the increase of investment spending has to be balanced with concerns on future debt sustainability—an issue particularly pressing in many developing countries that need large investment to achieve the targets of the 2030 Sustainable Development Agenda (see Box 1).35
79. If Ricardian Equivalence does not hold, the decrease in public saving associated with debt accumulation will not be fully compensated by higher private saving, and will lead to a lower stock of capital, resulting in higher interest rates and lower economic growth (Diamond, 1965; Blanchard, 1985). This is the classic crowding-out effect, which can also be obtained with a simple IS-LM model. Using the back-of-the envelope calculations of Elmendorf and Mankiw (1999), Panizza and Presbitero (2013) show that this effect is not quantitatively large.36 The crowding out effect of public debt could, however, become large if, in the presence of credit rationing and financial frictions, government debt tightens the credit constraints faced by private firms (Broner, Erce, Martin, and Ventura, 2014). The recent evidence of the European sovereign debt crisis has shown that the expansion of the share of government debt held by the banking sector in times of crisis crowds out private sector lending (Altavilla, Pagano and Simonelli 2017; Becker and Ivashina 2018).
80. High public debt can also have a negative effect on economic activity by increasing uncertainty about future tax rates, by leading to expectations of future confiscation, possibly through inflation and financial repression (Cochrane 2011), or by precipitating a financial crisis. There is, in fact, evidence that high levels of public debt could signal debt sustainability concerns and translate in higher sovereign yield spreads (Codogno et al., 2003; Laubach, 2009; Baum et al., 2013), which are then transmitted to the private sector. In fact, the debt overhang literature (Krugman, 1988; Sachs, 1989; Aguiar, Amador and Gopinath, 2009) suggests that there is a level of debt at which these growth effects are so large that debt relief would benefit both debtors and creditors.
81. Governments generally react to increasing public debt with austerity measures, running smaller deficits or larger surpluses (Bohn, 1998; Mendoza and Ostry 2008; Ghosh et al., 2013; Mauro et al., 2015). In this respect, high levels of public debt may also have a negative impact on growth as they could limit a country’s ability to conduct countercyclical policies and, possibly lead to self-defeating austerity policies, and thus increase output volatility and reduce economic growth. As the relationship between the level of debt and the ability of conduct countercyclical policies is also dependent on the composition of public debt (Hausmann and Panizza, 2011; De Grauwe, 2011), countries with different debt structures may start facing problems at very different levels of debt.
A. What Do the Data Say?
82. The rapid increase in public debt in the aftermath of the Global Financial Crisis sparked a large empirical literature on the growth effects of public debt.
83. An influential paper by Reinhart and Rogoff (2010a) uses data for 20 advanced economies over 1946–2009 to build a histogram that plots average GDP growth for different levels of debt and shows that average and median growth is substantially lower when public debt surpasses 90 percent of GDP. Figure 3 shows that this finding is robust to using more recent data from the newly available Global Debt Dataset (Mbaye et al., 2018). The Figure shows that average and median growth rates become smaller when moving to higher debt-to-GDP ratios. Average (median) growth declines from 3.7 percent in country-year pairs when the debt-to-GDP ratio is less than 30 percent to 2.6 (2.7) percent when the debt ratio is between 30 and 60 percent and further decreases to 1.2 (1.6) percent when debt surpasses 90 percent of GDP. However, these differences are smaller when looking at a large sample of 119 low- and middle-income countries, where average growth declines from 4.4 percent for low-debt countries to 2.6 percent in high-debt (above 90 percent) countries (Figure 4).
Government Debt and Growth, Selected Advanced Economies; 1960–2016
Citation: IMF Working Papers 2019, 101; 10.5089/9781498312103.001.A001
Source: Global Debt Dataset (Mbaye et al., 2018) and World Economic OutlookNotes: The sample includes 20 advanced economies as in Reinhart and Rogoff (2010, Figure 2): Australia, Austria, Belgium, Canada, Denmark, Finland, France, Germany, Greece, Ireland, Italy, Japan, Netherlands, New Zealand, Norway, Portugal, Spain, Sweden, the United Kingdom, and the United States. Data refer to central government debt, apart from the Netherlands, for which general government data have been used, because of data availability.
Government Debt and Growth, Selected Advanced Economies; 1960–2016
Citation: IMF Working Papers 2019, 101; 10.5089/9781498312103.001.A001
Source: Global Debt Dataset (Mbaye et al., 2018) and World Economic OutlookNotes: The sample includes 20 advanced economies as in Reinhart and Rogoff (2010, Figure 2): Australia, Austria, Belgium, Canada, Denmark, Finland, France, Germany, Greece, Ireland, Italy, Japan, Netherlands, New Zealand, Norway, Portugal, Spain, Sweden, the United Kingdom, and the United States. Data refer to central government debt, apart from the Netherlands, for which general government data have been used, because of data availability.Government Debt and Growth, Selected Advanced Economies; 1960–2016
Citation: IMF Working Papers 2019, 101; 10.5089/9781498312103.001.A001
Source: Global Debt Dataset (Mbaye et al., 2018) and World Economic OutlookNotes: The sample includes 20 advanced economies as in Reinhart and Rogoff (2010, Figure 2): Australia, Austria, Belgium, Canada, Denmark, Finland, France, Germany, Greece, Ireland, Italy, Japan, Netherlands, New Zealand, Norway, Portugal, Spain, Sweden, the United Kingdom, and the United States. Data refer to central government debt, apart from the Netherlands, for which general government data have been used, because of data availability.
Government Debt and Growth, Low and Middle Income Countries; 1960–2016
Citation: IMF Working Papers 2019, 101; 10.5089/9781498312103.001.A001
Source: Global Debt Dataset (Mbaye et al., 2018), World Development Indicators and World Economic Outlook.Notes: Data refer to central government debt. The sample includes 131 low- and middle-income countries. Data refer to central government debt.
Government Debt and Growth, Low and Middle Income Countries; 1960–2016
Citation: IMF Working Papers 2019, 101; 10.5089/9781498312103.001.A001
Source: Global Debt Dataset (Mbaye et al., 2018), World Development Indicators and World Economic Outlook.Notes: Data refer to central government debt. The sample includes 131 low- and middle-income countries. Data refer to central government debt.Government Debt and Growth, Low and Middle Income Countries; 1960–2016
Citation: IMF Working Papers 2019, 101; 10.5089/9781498312103.001.A001
Source: Global Debt Dataset (Mbaye et al., 2018), World Development Indicators and World Economic Outlook.Notes: Data refer to central government debt. The sample includes 131 low- and middle-income countries. Data refer to central government debt.84. Reinhart and Rogoff’s (2010a) article was followed by a large number of papers aimed at assessing whether the correlation between debt and growth was robust to controlling for other variables in a formal regression set-up, and to instrumenting public debt to assess its causal effect on economic growth. Another set of papers focuses on non-linearities allowing for non-arbitrary debt bracket.
85. By and large, there is strong evidence that public debt is negatively correlated with future economic growth. We corroborate the negative correlation between debt and growth by plotting current debt level and future growth and showing that, controlling for year- and country-fixed effects, there is a strong negative correlation between the debt-to-GDP ratio in year t and real GDP growth between t and t+5 (Figure 5). This negative correlation does not necessarily imply that high levels of debt cause lower growth. Indeed, the negative correlation between debt and growth could be driven by unobservable omitted variables that are jointly correlated with debt and growth, or simply pick up reverse causality. Slow economic growth, in fact, is an important factor explaining the rise of the public debt-to-GDP ratio (see, for instance, the Italian experience discussed in Box 2). This is not only because of the direct effect of growth on the denominator of the debt ratio, but also because the primary surplus depends on economic growth. In fact, absent any policy measure, low growth acts as constraint to public revenues, while expenditures increase in line with inflation, leading to a larger deficit and a rising debt (Mauro and Zilinsky, 2016).37 There is also a third factor that goes beyond the mechanical impact of growth on GDP and revenues. Mauro et al. (2015) show that, permanent growth slow down are often associated with increasing debt because policymakers often confuse changes in trend growth with temporary slowdowns and try to use expansionary fiscal policy to address a structural problem.
Government Debt and Subsequent GDP Growth, Selected Advanced Economies; 1960–2016
Citation: IMF Working Papers 2019, 101; 10.5089/9781498312103.001.A001
Source: Global Debt Dataset (Mbaye et al., 2018) and World Economic Outlook.Notes: A regression of the annual real GDP growth between t+5 and t against the ratio of general government debt over GDP at time t, controlling for year and country fixed effects, gives a coefficient on the debt variable of -0.016 (p-value of 0.001), meaning that 10 percent higher debt-to-GDP ratios are associated with 0.2 percent lower future growth over 5 years. To generate the binned scatterplot, starting from the sample of 19 OECD economies (data on general government for New Zealand are not available), the annual real GDP growth between t+5 and t (y-axis) and the ratio of general government debt over GDP at time t (x-axis) are regressed against year and country fixed effects. Then, the x-residuals are grouped into 50 equal-sized bins, then the chart plots, for each bin, the mean of the annual real GDP growth between t+5 and t, within each bin, holding the controls constant. The red line is the linear fit of the OLS regression of the y-residuals on the x-residuals. The number of observations is 923.
Government Debt and Subsequent GDP Growth, Selected Advanced Economies; 1960–2016
Citation: IMF Working Papers 2019, 101; 10.5089/9781498312103.001.A001
Source: Global Debt Dataset (Mbaye et al., 2018) and World Economic Outlook.Notes: A regression of the annual real GDP growth between t+5 and t against the ratio of general government debt over GDP at time t, controlling for year and country fixed effects, gives a coefficient on the debt variable of -0.016 (p-value of 0.001), meaning that 10 percent higher debt-to-GDP ratios are associated with 0.2 percent lower future growth over 5 years. To generate the binned scatterplot, starting from the sample of 19 OECD economies (data on general government for New Zealand are not available), the annual real GDP growth between t+5 and t (y-axis) and the ratio of general government debt over GDP at time t (x-axis) are regressed against year and country fixed effects. Then, the x-residuals are grouped into 50 equal-sized bins, then the chart plots, for each bin, the mean of the annual real GDP growth between t+5 and t, within each bin, holding the controls constant. The red line is the linear fit of the OLS regression of the y-residuals on the x-residuals. The number of observations is 923.Government Debt and Subsequent GDP Growth, Selected Advanced Economies; 1960–2016
Citation: IMF Working Papers 2019, 101; 10.5089/9781498312103.001.A001
Source: Global Debt Dataset (Mbaye et al., 2018) and World Economic Outlook.Notes: A regression of the annual real GDP growth between t+5 and t against the ratio of general government debt over GDP at time t, controlling for year and country fixed effects, gives a coefficient on the debt variable of -0.016 (p-value of 0.001), meaning that 10 percent higher debt-to-GDP ratios are associated with 0.2 percent lower future growth over 5 years. To generate the binned scatterplot, starting from the sample of 19 OECD economies (data on general government for New Zealand are not available), the annual real GDP growth between t+5 and t (y-axis) and the ratio of general government debt over GDP at time t (x-axis) are regressed against year and country fixed effects. Then, the x-residuals are grouped into 50 equal-sized bins, then the chart plots, for each bin, the mean of the annual real GDP growth between t+5 and t, within each bin, holding the controls constant. The red line is the linear fit of the OLS regression of the y-residuals on the x-residuals. The number of observations is 923.The Evolution of Italian Public Debt After World War II
Italy has the fourth largest stock of public debt in the world, the second highest debt-to- GDP ratio in the G7 group of advanced economies, and the highest debt service ratio in the G7.
Italian public debt stood at 74 percent of GDP at the end of World War II in 1945, but dropped to 24 percent of GDP by 1947.1/ The main driver of this rapid debt reduction was high inflation (20 percent in 1946 and 62 percent in 1947) and the corresponding negative real interest rates (the stock of debt still carried single digit interest rates). From this point on, it is possible to identify 7 phases in the behavior of the Italian public debt: (i) a moderate increase (to 34 percent of GDP) during the post war reconstruction period; (ii) a slight decrease (to 27 percent of GDP) over 1955–1964; (iii) a rapid increase to (to 56 percent of GDP) over 1964–1975; (iv) a period of relative stability in the second half of the 1970s; (v) a rapid increase over 1980–1994, with debt peaking at 120 percent of GDP in 1994; (vi) a consolidation phase over 1995–2007, with debt bottoming at 103 of GDP in 2004; and (vii) a rapid increase after 2008, with debt surpassing 130 percent of GDP in 2016.
These different phases can be explained by different economic and political events. The increase in debt during the reconstruction period was driven by massive public investments, which were needed to rebuild and modernize Italian infrastructures (by the early 1960s, public investment still accounted for more than 20 percent of total public expenditure). Over 1955–1964, debt reduction was driven by a combination of rapid economic growth and small overall budget deficits (composed of small primary surpluses that almost fully balanced interest payments; in the post-war period Italy never ran an overall budget surplus). In the second half of the 1960s, Italy started running increasingly larger primary deficits, hovering at around 2 percent GDP in the second half of the 1960s and going well above 4 percent of GDP in the 1970s. These large budget deficits are partly explained by the political economy theories illustrated in Section 3 as they were associated to a period of often short-lived coalition governments, a strengthening of the opposition Communist Party (which in late 1960s and early 1980s was receiving more than 30 percent of Italian votes), and a period of political and labor tensions that culminated in the “Hot Autumn” of 1969–70.2/
The interesting fact is that while primary deficits led to a debt explosion in the second half of the 1960s and early 1970s, even larger primary deficits had almost no impact on debt in the second half of the 1970s. The explanation has to do with the conduct of monetary policy which, by the late 1970s was mostly driven by the need to finance growing deficits. Debt monetization led to high inflation and, with the help of some financial repression, negative real interest rates, which kept debt under control over 1973–1980.3/
The situation changed when, in 1981, the “divorce” between Bank of Italy and the Treasury prevented the central bank from monetizing the deficit. The divorce led to a rapid reduction of inflation and a sudden increase of real interest rates.4/ The conquest of fiscal dominance, however, was not followed by fiscal reforms, and Italy kept running primary deficits until 1991. The combination of prudent monetary policy and large primary deficits led to the explosion of public debt over the 1980s and early 1990s. Starting from 1991, Italy began running primary surpluses. However, with high real interest rates debt kept growing until 1994. In the first half of the 1990s, interested payments absorbed more than 20 percent of total public outlays (in 1992 Italy ran a primary surplus of 1.8 percent GDP but the overall budget deficit was above 10 percent of GDP).
Starting in 1994, a combination of substantial primary surpluses and decreasing real rates lead to a process of debt reduction. This process of fiscal consolidation, driven by higher revenues (revenues reached 47 percent of GDP in 1997, partly due to special “below the line” operations) and unchanged primary expenditure (at about 41 percent of GDP), slowed down over 2000–2005. Primary surpluses went from the 4–6 percent range of the late 1990s to a 1–3 percent range, as a consequence, the debt-to-GDP ratio started growing again in 2005–2006. There was a brief moment of consolidation in 2007 immediately interrupted by the explosion of the global financial crisis in 2008/2009. The first half of the new millennium was thus a missed opportunity for further reducing public debt, and Italy entered the global financial crisis with a level of debt well above 100 percent of GDP.
The post global financial crisis period was characterized by fiscal restraint (with Italy running a small primary deficit in 2009 and 2010 and primary surpluses well above its European counterparts in the remaining years), but with a contracting economy, which lead to a rapid increase in the debt to GDP ratio. During the crisis, the high level of debt associated with past fiscal profligacy constrained Italy’s policy response and possibly contributed to magnifying the crisis. In this sense, the high level of inherited debt may have a led to suboptimal macroeconomic policies with a negative effect on Italy’s GDP growth during the crisis—and, correspondingly, worsening debt dynamics.
1/ All the data refer to gross debt. In Italy the difference between gross and net debt is not very large. For instance, WEO data for 2016 report a gross debt of 131 percent of GDP and a net debt of 119 percent of GDP.2/ The term “Hot Autumn” denotes a series of large strikes in the main industrial cities of Northern Italy.3/ Consumer price inflation was almost always above 15 percent over 1974–1982 (it was 12 percent in 1978) and peaked at 21 percent in 19804/ Inflation went from 21 percent in 1980, to 9 percent in 1985, 5 percent in 1987, and 2 percent in 199786. Thus, establishing whether debt has a causal effect on growth requires an instrumental variable or a natural experiment that allows the researcher to isolate exogenous changes in public debt. In the presence of persistent variables like the debt-to-GDP ratio, the standard approach of using of past values of the variables of interest as instruments does not solve the identification problem (Bellemare et al., 2017; Reed, 2015). Panizza and Presbitero (2014) propose a strategy that uses valuation effects brought about by the presence of foreign currency debt as an exogenous driver of the change in public debt and find that public debt has no effect on future growth. One problem with this approach is that in their sample of advanced economies the share of foreign currency-denominated debt is relatively small, and hence the instrument is not very strong. Another way to achieve identification is to move from macro to micro data. Huang et al. (2017, 2018) match firm-level balance sheets with data on either public debt across a sample of 69 countries or local government debt across 270 Chinese cities to show that government debt tightens financing constraints for private sector manufacturing firms. In a similar vein, Croce et al. (2019) identify a different channel through which debt can affect productivity and growth showing that, in the US, higher government debt increases the cost of capital and has a negative effect on investment by R&D-intensive firms. There is, however, a tradeoff between identification and the ability of assessing the macroeconomic effects of debt accumulation. While firm-level analysis allows to precisely test one channel through which debt may have a negative effect on growth, they “hide” the potential macroeconomic links between debt and growth. For instance, it is possible that higher levels of debt increase investment for all industries and firms considered by Huang et al. (2017, 2018), but that investment increases less for credit-constrained firms.
87. Besides studying the average correlation between debt and growth, the economics literature also seeks to identify possible non-linearities and threshold effects. Some analyses indicate that the debt trajectory can have more important consequences for economic growth than the level of debt-to-GDP itself (Pescatori et al., 2014; Chudik et al. 2017), in line with recent evidence on how public debt could affect debt sustainability and market access (Bassanetti et al., 2018).
88. Moving to the presence of debt thresholds, the notion that there is a non-linearity in the debt-growth relationship and that this non-linearity is at a specific value of the debt-to-GDP ratio—often taken to be 90 percent—has become popular.38 However, assessing non-linearities is complicated by lack of statistical power due to the limited number of observations above the relevant threshold, and it is possible that the results of the literature are driven by the imposition of some parametric approach and on a few outliers.39 Moreover, this literature imposes common coefficients and thresholds across countries, while the data suggest that there is substantial heterogeneity especially when looking at larger samples, which pool together developing and emerging economies as well (Eberhardt and Presbitero, 2015; Chudik et al. 2017). Consider, for instance, Figure 6, which plots the outcome of a non-parametric regression based on a sample of 20 advanced economies over the period 1960–2016 and shows that (i) the average negative correlation between debt and future (5 years ahead) growth hides a large degree of heterogeneity across countries, and (ii) while the relationship between debt and growth is nonlinear there is no common threshold beyond which an increase in debt is associated with a growth slowdown.
Non-linearities and Heterogeneity in the Debt-Growth Relationship
Citation: IMF Working Papers 2019, 101; 10.5089/9781498312103.001.A001
Source: Global Debt Dataset (Mbaye et al., 2018) and World Economic Outlook.Notes: The solid black line plots smoothed values of the locally weighted regression of the annual real GDP growth between t+5 and t against the ratio of general government debt over GDP at time t for the whole sample. The thin lines plot the same smoothed values for single countries. The histogram shows the density distribution of the ratio of general government debt over GDP (x-axis). The sample includes 20 advanced economies as in Reinhart and Rogoff (2010, Figure 2): Australia, Austria, Belgium, Canada, Denmark, Finland, France, Germany, Greece, Ireland, Italy, Japan, Netherlands, New Zealand, Norway, Portugal, Spain, Sweden, the United Kingdom, and the United States. Data refer to central government debt, apart from the Netherlands, for which general government data have been used, because of data availability.
Non-linearities and Heterogeneity in the Debt-Growth Relationship
Citation: IMF Working Papers 2019, 101; 10.5089/9781498312103.001.A001
Source: Global Debt Dataset (Mbaye et al., 2018) and World Economic Outlook.Notes: The solid black line plots smoothed values of the locally weighted regression of the annual real GDP growth between t+5 and t against the ratio of general government debt over GDP at time t for the whole sample. The thin lines plot the same smoothed values for single countries. The histogram shows the density distribution of the ratio of general government debt over GDP (x-axis). The sample includes 20 advanced economies as in Reinhart and Rogoff (2010, Figure 2): Australia, Austria, Belgium, Canada, Denmark, Finland, France, Germany, Greece, Ireland, Italy, Japan, Netherlands, New Zealand, Norway, Portugal, Spain, Sweden, the United Kingdom, and the United States. Data refer to central government debt, apart from the Netherlands, for which general government data have been used, because of data availability.Non-linearities and Heterogeneity in the Debt-Growth Relationship
Citation: IMF Working Papers 2019, 101; 10.5089/9781498312103.001.A001
Source: Global Debt Dataset (Mbaye et al., 2018) and World Economic Outlook.Notes: The solid black line plots smoothed values of the locally weighted regression of the annual real GDP growth between t+5 and t against the ratio of general government debt over GDP at time t for the whole sample. The thin lines plot the same smoothed values for single countries. The histogram shows the density distribution of the ratio of general government debt over GDP (x-axis). The sample includes 20 advanced economies as in Reinhart and Rogoff (2010, Figure 2): Australia, Austria, Belgium, Canada, Denmark, Finland, France, Germany, Greece, Ireland, Italy, Japan, Netherlands, New Zealand, Norway, Portugal, Spain, Sweden, the United Kingdom, and the United States. Data refer to central government debt, apart from the Netherlands, for which general government data have been used, because of data availability.89. One reason for the presence of country-specific thresholds is that the level at which public debt becomes “too high” must depend on country characteristics. For example, in the context of sovereign default, Reinhart, Rogoff and Savastano (2003) classified countries into clubs and “debt intolerance” regions, which depend not only on borrowers’ debt levels, but also on their credit and inflation history. Likewise, Ghosh et al. (2013), in calculating fiscal space, find that governments’ debt-sustainability thresholds depend on their historical track record of fiscal adjustment in response to rising public debt. Alternatively, Eichengreen, Hausmann, and Panizza (2005) emphasized the role of debt composition. In the debt and growth literature, Kourtellos et al. (2013) explicitly modeled the possibility of different regimes depending on a large set of country characteristics and find that only when institutions are below a certain level higher debt translates into lower GDP growth. As countries with poor institutions also have higher debt levels, these results provide a mechanism to interpret (and are consistent with) the general finding of a negative relationship between debt and growth. Specifically, countries with low-quality institutions may be more inclined to the political budget cycle and less able to control overborrowing. In addition, those countries could have a higher propensity to finance government consumption rather than productive investment, leading to higher debt and lower growth.
90. Other authors have looked at the dynamics between debt and growth from an historical perspective. Esteve and Tamarit (2018) focus on the Spanish economy for the period 1851–2013 and find some support for a negative relationship between public debt and growth, but no clear evidence of a debt threshold. Balassone et al. (2013) consider the experience of Italy since its unification in 1861 and find that when debt exceeds 100 percent of GDP its negative effect on growth becomes stronger. Eberhardt (2017) challenges this conclusion on grounds that time series analysis (e.g., cointegration) does not imply causation and is not suited to look at non-linearities. Adopting a more flexible framework and data over more than two centuries for Great Britain, Japan, Sweden, and the United States, Eberhardt (2017) finds no evidence for any long-run non-linear relationship between debt and growth.
91. Overall, our reading of the empirical literature is that, at least in advanced economies, there is a negative correlation between public debt and subsequent economic growth but no convincing evidence of causality: high debt and low growth may just reflect a weak macroeconomic framework, which is driving both aggregates. Moreover, cross-country averages hide a more complex reality, as the debt and growth relationship is driven by a number of factors that differ across countries and no clear evidence of a common tipping point beyond which additional debt has a negative effect on growth.
92. Six years after Reinhart, Reinhart and Rogoff (2012) we still agree with them that the “endogeneity conundrum has not been fully resolved” (p. 80).
93. Finally, it bears emphasizing that even if is true (in a causal sense) that “debt is bad for growth” it does not necessarily follow that governments should pay down the existing debt (Ostry, Ghosh, and Espinoza, 2015). In terms of social welfare (i.e., the distortionary cost), it may be more costly to pay down the debt than to live with it (provided, of course, the government does not face a funding crisis). In steady-state, this result is follows directly from tax-smoothing. Unless taxes are set to just service the debt indefinitely, they will either have to be increased in order to maintain sustainability against a growing debt, or they will have to be decreased once the debt has been repaid—either violates the principle of smoothing taxes to minimize the distortionary costs. Interestingly, the result also holds out of steady-state—at least for an important class of utility functions (i.e., iso-elastic). This is because, even though the presence of distortionary taxes implies wedges between private and social marginal products and rates of substitution, the market interest rate equals the discount rate of a benevolent government (i.e., that seeks to maximize the representative agent’s utility).40 The government can choose to pay down $1 of domestic debt today at a certain distortionary cost. Or it can wait till tomorrow, when the debt and the cost will have grown by (1+r), the market interest rate. But the government discounts the future at precisely (1+r), so it is indifferent between paying down the debt today or tomorrow. Since the same argument holds across all periods, the steady-state result—that it is optimal to just live with the inherited debt—obtains even out of steady state.
B. Not All Debts Are Equal
94. One of the reasons why it is difficult to identify common patterns and pin down the causal effect of debt on growth is that not all debts are equal and factors such as what the debt was used for, who holds government debt, its currency composition and its maturity are key elements that can affect fiscal vulnerabilities, the use of debt for tax smoothing purposes, and how government and private agents could react to future changes in debt. Unfortunately, applied economists are not usually able to observe all these characteristics, but only the level of government debt.41 In other words, treating debt as a black box and imposing the restriction that any given level of debt has the same consequence on economic growth, regardless of its structure is too simplistic. We need to factor into the empirical analysis all the other elements that matter (Eberhardt and Presbitero, 2015; Chudik et al., 2017).
95. Even in the studies where debt is allowed to affect countries in different ways, the lack of data availability on its structure precludes analysis—and hence a clear understanding—of what is driving the heterogeneity in the results. The key issue, especially from a policy perspective, is not just to acknowledge that the economic consequences of borrowing are different across countries, but to understand which factors explain why a given debt level could be a constraint in one country but not in another. For instance, for any given debt level, having a larger exposure to foreign creditors—the share of public debt held by non-residents is traditionally very low for Japan (around 5–7 percent), while it was close to 40 percent in Italy and even higher in Ireland (Abbas et al., 2014)—or a shorter maturity structure, increases debt vulnerabilities, hampering growth (of course debt structure is itself endogenous, adding another layer of complication ever if the data were available).
96. Similarly, looking at gross debt measures, as is normally done—including in this paper—may be misleading, given that gross debt ignores financial assets held by the government (as well as the share of debt held directly by the government through cross-holdings). In some countries, this is particularly large, making their net stock of debt significantly smaller than the gross debt used in the empirical analyses. On the other hand, implicit liabilities, such as pensions, and debt by local governments and state-owned enterprises often results in much larger public sector debt stocks than the official debt numbers would suggest. However, capacity constraints in debt management offices and lack of data makes it difficult to have a common definition of net and implicit debts, which are comparable across countries (Panizza and Presbitero, 2013).
V. Concluding Remarks
97. Governments issue debt for a variety of reasons—both good and bad. Among the good reasons are intertemporal tax-smoothing, fiscal stimulus during economic downturns, and asset management, including providing financial markets with safe assets. While such motives can explain some of the increases in public debt—in particular, after wars or major financial crises—they cannot plausibly account for all of the observed changes. The correlation between public investment and public borrowing—supposedly a major non-wartime motivation for issuing debt—is surprisingly weak. Indeed, the behavior of governments is sometimes quite at odds with these theories. A notable example is the buildup of public debt in many advanced economies during the early 2000s, when the world economy was booming, and the looming prospect of aging-related costs should have spurred public saving.
98. Counter-cyclical fiscal policies with implementation delays and forecast biases might be part of the explanation for the upward trend in public debt in many advanced economies. But a full accounting needs to go beyond purely economic rationales and consider social, political, and institutional factors that might be at play. Politicians pursuing their own self-interest and seeking to maximize their chances of re-election may engage in a political business cycle that results in debt rising over time. Strategic manipulation whereby the party in power seeks to circumscribe its (possible) successor’s ability to spend public funds by deliberately running up public debt will likewise result in a positive debt bias. And common pool problems, which result in the private benefit of an additional unit of spending exceeding the social marginal cost of funding this extra unit of expenditure, provide a third political economy explanation. More generally, where the “safety-valve” of inflation is unavailable (e.g., under a currency board regime or membership in a monetary union), competing demands by different socio-economic groups is often (temporarily) resolved through rising public debt.
99. But why does overborrowing matter? And what can be done about it? Other chapters in in Abbas, Pienkowski and Rogoff (2019) explore some of the consequences of excessive government borrowing—including debt sustainability problems and possible crises. Even in the absence of crises, however, public debt can be costly. In welfare terms, the cost of public debt is the present discounted value of distortions associated with the taxes necessary to service that debt. Empirically, there is a negative relationship between public debt and output growth. The jury is still out on whether that relationship is causal—higher levels of public debt impeding growth—and in reality, the answer must depend on what the debt was used to finance, how it is expected to be repaid or serviced, and a host of other country-specific factors.
100. As to measures that democracies can take to limit overborrowing, the literature has identified three key avenues: electoral systems, fiscal rules, and budgetary institutions. While their effectiveness will depend on country circumstances, all imply some trade-off between the flexibility to respond to shocks and to issue debt for good economic reasons and the need to discipline policy makers from borrowing excessively.
References
Abbas, SM Ali, and Jakob E. Christensen, 2010, The role of domestic debt markets in economic growth: An empirical investigation for low-income countries and emerging markets, IMF Staff Papers 57, 209–255.
Abbas, SM Ali, Laura Blattner, Mark De Broeck, Asmaa El-Ganainy, and Malin Hu, 2014, Sovereign Debt Composition in Advanced Economies: A Historical Perspective, IMF Working Paper, No. 14/162.
Abbas, SM Ali, Alex Pienkowski, and Kenneth S. Rogoff, (eds), 2014, Sovereign Debt: A Guide for Economists and Practitioners, Oxford: Oxford University Press
Abel, Andrew B, N Gregory Mankiw, Lawrence H Summers, and Riachard J Zeckhauser, 1989, Assessing dynamic efficiency: Theory and evidence, The Review of Economic Studies 56, 1–19.
Abiad, Abdul, Davide Furceri, and Petia Topalova, 2016, The macroeconomic effects of public investment: Evidence from advanced economies, Journal of Macroeconomics, 50, 224–240.
Aghion, Philippe and Patrick Bolton, 1990, Government Domestic Debt and the Risk of Default: A Political- Economic Model of the Strategic Role of Debt, in: Rudi Dornbusch and Mario Draghi, Government Domestic Debt and the Risk of Default: A Political-Economic Model of the Strategic Role of Debt, Cambridge: The MIT Press.
Aghion, Philippe, Ioana Marinescu, Ricardo J. Caballero, and Anil K. Kashyap, 2007, Cyclical Budgetary Policy and Economic Growth: What Do We Learn from OECD Panel Data? [with Comments and Discussion], NBER Macroeconomics annual 22, 251–297.
Aguiar, Mark, and Manuel Amador, 2011, Growth in the Shadow of Expropriation, Quarterly Journal of Economics, 126, 651–697.
Aguiar, Mark, Manuel Amador, and Gita Gopinath, 2009, Investment Cycles and Sovereign Debt Overhang, The Review of Economic Studies, 76, 1–31.
Aiyagari, S. Rao, Albert Marcet, Thomas J. Sargent, and Juha Seppälä, 2002, Optimal taxation without state-contingent debt, Journal of Political Economy 110, 1220–1254.
Alesina, Alberto, and Alex Cukierman, 1990. The Politics of Ambiguity, The Quarterly Journal of Economics, 105 (4), 829–850.
Alesina, Alberto, and Andrea Passalacqua, 2016, The Political Economy of Government Debt, in: John Taylor and Harald Uhlig (eds.) The Handbook of Macroeconomics, Amsterdam: Elsevier, 2599–2651.
Alesina, Alberto and Roberto Perotti, 1995. Fiscal expansions and adjustments in OECD countries, Economic Policy 10, 205–248.
Alesina, Alberto, and Guido Tabellini, 1990, A positive theory of fiscal deficits and government debt, Review of Economic Studies 57 (3), 403–414.
Alesina Alberto and Guido Tabellini, 1992, Positive and normative theories of public debt and inflation in historical perspective, European Economic Review 36, 337–44.
Alesina, Alberto, and Roberto Perotti, 1996, Fiscal Discipline and the Budget Process, The American Economic Review P&P, 86 (2), 401–407.
Alesina, Alberto, Filipe R. Campante, and Guido Tabellini, 2008, Why is fiscal policy often procyclical?, Journal of the European Economic Association 6, 1006–1036.
Alesina, Alberto, Ricardo Hausmann, Rudolf Hommes, Ernesto Stein, 1999, Budget Institutions and Fiscal Performance in Latin America, Journal of Development Economics, 59 (2), 253–273.
Alt, James E., and David D. Lassen, 2006, Fiscal transparency, political parties, and debt in OECD countries, European Economic Review, 50, 1403–1439.
Altavilla, Carlo, Marco Pagano and Saverio Simonelli, 2017, Bank Exposures and Sovereign Stress Transmission, Review of Finance, 21 (6), 2103–2139.
Amaglobeli, David, Nicolas End, Mariusz Jarmuzek, and Geremia Palomba, 2017, The fiscal costs of systemic banking crises, International Finance, 20 (1), 2–25.
Angeletos, George-Marios, 2002, Fiscal policy with noncontingent debt and the optimal maturity structure, The Quarterly Journal of Economics 117, 1105–1131.
Ash, Michael, Deepankar Basu, and Arindrajit Dube, 2017, Public debt and growth: An assessment of key findings on causality and thresholds, PERI Working Paper, No. 2017–10.
Auerbach, Alan J., and Yuriy Gorodnichenko, 2013, Fiscal Multipliers in Recession and Expansion, in: Alberto Alesina and Francesco Giavazzi (eds.) Fiscal Policy and the Financial Crisis, Chicago: University of Chicago Press.
Azzimonti, Marina, Marco Battaglini, and Stephen Coate, 2016, The costs and benefits of balanced budget rules: Lessons from a political economy model of fiscal policy, Journal of Public Economics, 136(C), 45–61.
Bacchiocchi, Emanuele, Elisa Borghi, and Alessandro Missale, 2011. Public investment under fiscal constraints. Fiscal Studies, 32 (1), 11–42.
Balassone, Fabrizio, Maura Francese, and Angelo Pace, 2013, Public debt and economic growth: Italy’s first 150 years, in Gianni Toniolo (ed.), The Oxford Handbook of the Italian Economy Since Unification, Oxford: Oxford University Press.
Banerji, Angana, Valerio Crispolti, Era Dabla-Norris, Christian H. Ebeke, Davide Furceri, Takuji Komatsuzaki, and Tigran Poghosyan, 2017, Labor and product market reforms in advanced economies: fiscal costs, gains, and support, International Monetary Fund, Washington DC.
Baron, David, and John Ferejohn, 1989, Bargaining in Legislatures, The American Political Science Review, 83 (4), 1181–1206.
Barro, Robert J., 1979, On the Determination of the Public Debt, Journal of Political Economy 87, 940.
Barro, Robert J., 1995, Optimal debt management, National Bureau of Economic Research.
Bassanetti, Antonio, Carlo Cottarelli, and Andrea F. Presbitero, 2018, Lost and found: Market access and public debt dynamics, Oxford Economic Papers, 71 (2), 445–471.
Battaglini, Marco, 2010, The Political Economy of Public Debt, Annual Review of Economics, 3 (1), 161–189.
Battaglini, Marco, 2014, A Dynamic Theory of Electoral Competition, Theoretical Economics, 9, 515–554.
Battaglini, Marco, and Stephen Coate, 2008, A Dynamic Theory of Public Spending, Taxation, and Debt, The American Economic Review, 98 (1), 201–236.
Becker, Bo, and Victoria Ivashina, 2018, Financial Repression in the European Debt Crisis, Review of Finance, 22(1): 83–115.
Beetsma, Roel MWJ, and Henrik Jensen, 2005, Monetary and fiscal policy interactions in a Micro-Founded Model of a Monetary Union, Journal of international Economics 67, 320–352.
Beetsma, Roel, Xavier Debrun, and Randolph Sloof, 2017, The Political Economy of Fiscal Transparency and Independent Fiscal Councils, IMF Working Paper, No. 17/195
Bellemare, Marc F., Takaaki Masaki, and Thomas B. Pepinsky, 2017, Lagged explanatory variables and the estimation of causal effect, The Journal of Politics, 79 (3), 949–963.
Berck, Peter, and Jonathan Lipow, 2011, Beyond Tax Smoothing, unpublished working paper.
Bergman, Michael, Michael Hutchison, and Svend Hougaard Jensen, 2016, Promoting sustainable public finances in the European Union: the role of fiscal rules and government efficiency, European Journal of Political Economy, 44, 1–19.
Bhandari, Anmol, David Evans, Mikhail Golosov, and Thomas J. Sargent, 2016, Fiscal policy and debt management with incomplete markets, The Quarterly Journal of Economics 132, 617–663.
Blanchard, Olivier J., 1985, Debt, deficits, and finite horizons, Journal of Political Economy 93, 223–247.
Blanchard, Olivier J., 1993, Suggestions for a New Set of Fiscal lndicators, The Political Economy of Government Debt. H. A. A. Verbon and F. A. A. M. Van Winden. (Amsterdam: Elsevier Science).
Blanchard, Olivier J., and Lawrence Summers, 2017, Rethinking stabilization policy: Back to the future, Rethinking Macroeconomic Policy Conference, Peterson Institute for International Economics, October.
Blanchard, Olivier, and Jordi Gali, 2010, Labor markets and monetary policy: A New Keynesian model with unemployment, American Economic Journal: Macroeconomics, 2, 1–30.
Blanchard, Olivier J., 2019, Public Debt and Low Interest Rates. AEA Presidential Address.
Bohn, Henning, 1990, Tax smoothing with financial instruments, The American Economic Review, 1217–1230.
Bohn, Henning, 1998, The behavior of US public debt and deficits, The Quarterly Journal of Economics, 113 (3), 949–963.
Bolton, Patrick and Olivier Jeanne, 2007, Structuring and Restructuring Sovereign Debt: The Role of a Bankruptcy Regime, Journal of Political Economy, 115 (6), 901–924.
Borensztein, Eduardo and Paolo Mauro, 2004. The case for GDP-indexed bonds, Economic Policy, 19, 165–216.
Borensztein, Eduardo, Eduardo Levy Yeyati and Ugo Panizza, 2006. Living with Debt, Harvard University Press for the IDB.
Bova, Elva, Marta Ruiz-Arranz, Frederik Toscani, and H. Elif Ture, 2016. The Fiscal Costs of Contingent Liabilities: A New Dataset, IMF Working Paper 16/14.
Braun, Matıas, et al., 2008. Development of the Chilean corporate bond market. In: Eduardo Borensztein, Kevin Cowan, Barry Eichengreen and Ugo Panizza, (eds.) Bond Markets in Latin America: On the Verge of a Big Bang: 151–184. Cambridge, MIT Press
Broner, Fernando, Aitor Erce, Alberto Martin, and Jaume Ventura, 2014, Sovereign debt markets in turbulent times: Creditor discrimination and crowding-out effects, Journal of Monetary Economics, 61(C), 114–142.
Brunnermeier, Markus K., Sam Langfield, Marco Pagano, Ricardo Reis, Stijn Van Nieuwerburgh, and Dimitri Vayanos, 2017, ESBies: Safety in the tranches, Economic Policy 32, 175–219.
Buchanan, James and Gordon Tullock, 1962. The Calculus of Consent: Logical Foundations of a Constitutional Democracy, University of Michigan Press
Buchanan, James, and Richard Wagner, 1977, The Political Legacy of Lord Keynes. New York: Academic Press.
Buera, Francisco, and Juan Pablo Nicolini, 2004, Optimal maturity of government debt without state contingent bonds, Journal of Monetary Economics, 51, 531–554.
Caballero, Ricardo J., Emmanuel Farhi, and Pierre-Olivier Gourinchas, 2008, An equilibrium model of “global imbalances” and low interest rates, The American Economic Review 98, 358–93.
Cafiso, Gianluca, 2012, Debt Developments and Fiscal Adjustment in the EU, Intereconomics, 47 (1), 61–72.
Calvo, Guillermo, 1988, Servicing the Public Debt: The Role of Expectations, The American Economic Review, 78 (4), 647–661.
Campos, Camila, Dany Jaimovich, and Ugo Panizza, 2006, The unexplained part of public debt, Emerging Markets Review, 7, 228–243.
Caselli, Francesca, and Philippe Wingender, 2018, Bunching at 3 percent: The Maastricht fiscal criterion and government deficits, IMF Working Paper, No. 18/182.
Chan, Louis Kuo Chi, 1983, Uncertainty and the neutrality of government financing policy, Journal of Monetary Economics 11, 351–372.
Chari, Varadarajan V., Lawrence J. Christiano, and Patrick J. Kehoe, 1994, Optimal fiscal policy in a business cycle model, Journal of Political Economy, 102, 617–652.
Checherita-Westphal, Cristina, and Philipp 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, Alexander, Kamiar Mohaddes, M. Hashem Pesaran, and Mehdi Raissi, 2017, Is there a debt-threshold effect on output growth?, The Review of Economics and Statistics, 99 (1), 135–150.
Cochrane, John H., 2011, Understanding policy in the great recession: Some unpleasant fiscal arithmetic, European Economic Review, 55 (1), 2–30.
Codogno, Lorenzo, Carlo Favero, and Alessandro Missale, 2003, Yield spreads on EMU government bonds, Economic Policy, 18, 503–532
Croce, Mariano M., Thien T. Nguyen, Steve Raymond, and Lukas Schmid, 2018, Government Debt and the Returns to Innovation, Journal of Financial Economics, forthcoming.
Cukierman, Alex and Allan Meltzer, 1989, A Political Theory of Government Debt and Decits in a Neo-Ricardian Framework, American Economic Review, 79, 713–732.
Dabla-Norris, Era, Richard Allen, Luis-Felipe Zanna, Tej Prakash, Eteri Kvintradze, Victor Lledo, Irene Yackovlev, and Sophia Gollwitzer, 2010, Budget Institutions and Fiscal Performance in Low-Income Countries, IMF Working Paper, No. 10/80.
Debrun, Xavier, Laurent Moulin, Alessandro Turrini, Joaquim Ayuso-i-Casals and Manmohan Kumar, 2008, Tied to the mast? National fiscal rules in the European Union, Economic Policy, 23 (4), 297–362.
DeLong, J. Bradford, and Lawrence H. Summers, 2012, Fiscal Policy in a Depressed Economy [with Comments and Discussion], Brookings Papers on Economic Activity, 233–297.
De Grauwe Paul, 2011, The Governance of a Fragile Eurozone, CEPS Working Documents, Centre for European Policy Studies.
Diamond, Peter A., 1965, National debt in a neoclassical growth model, The American Economic Review, 55 (5), 1126–1150.
Eberhardt, Markus and Andrea F. Presbitero, 2015, Public Debt and Growth: Heterogeneity and Non-linearity, Journal of International Economics, 97 (1), 45–58.
Eberhardt, Markus, 2017, Nonlinearities in the relationship between debt and growth: (no) evidence from over two centuries, Macroeconomic Dynamics, forthcoming.
Égert, Balázs, 2012, Fiscal Policy Reaction to the Cycle in the OECD: Pro-or Countercyclical?, CESifo Working Paper Series, No. 3777.
Eggertsson, Gauti B., and Michael Woodford, 2009, Optimal monetary and fiscal policy in a liquidity trap, in: Richard H. Clarida, Jeffrey Frankel, Francesco Giavazzi and Kenneth D. West (eds.) NBER International Seminar on Macroeconomics 2004, Cambridge: The MIT Press.
Eichengreen, Barry, and Ugo Panizza, 2016, A surplus of ambition: can Europe rely on large primary surpluses to solve its debt problem?, Economic Policy, 31 (85), 5–49.
Eichengreen, Barry, Ricardo Hausmann, and Ugo Panizza, 2005, The pain of original sin, in: Barry Eichengreen and Ricardo Hausmann (eds.) Other People’s Money: Debt Denomination and Financial Instability in Emerging Market Economies, Chicago: University of Chicago Press.
Eichengreen, Barry, Robert Feldman, Jeffrey Liebman, Jürgen von Hagen, and Charles Wyplosz, 2011, Public Debts: Nuts, Bolts and Worries, Geneva Reports on the World Economy, CEPR, London.
Elmendorf, Douglas W., and Gregory N. Mankiw, 1999, Government debt, in: John B. Taylor and Michael Woodford (eds.) Handbook of Macroeconomics, Elsevier, Chapter 25, 1615–1669
Esteve, Vicente, and Cecilio Tamarit, 2018, Public debt and economic growth in Spain, 1951–2013, Cliometrica, 12, 219–249.
Eyraud, Luc, Xavier Debrun, Andrew Hodge, Victor Duarte Lledo, and Catherine A Pattillo, 2018. Second-Generation Fiscal Rules: Balancing Simplicity, Flexibility, and Enforceability. Staff Discussion Notes, No. 18/04.
Fatás, Antonio, 2009, The Effectiveness of Automatic Stabilizers, Workshop on Fiscal Policy, IMF, Jun
Fatás, Antonio, 2018, Fiscal policy, potential output and the shifting goalposts, (Manuscript).
Fatás, Antonio, and Ilian Mihov, 2010, The euro and fiscal policy, Europe and the Euro, Chicago: University of Chicago Press.
Fatás, Antonio, and Ilian Mihov, 2012, Fiscal Policy as a Stabilization Tool, The BE Journal of Macroeconomics, 12, 1–68.
Fernandez, Raquel, and Dani Rodrik, 1991, Resistance to reform: Status quo bias in the presence of individual-specific uncertainty, The American Economic Review, 81, 1146–1155.
Frankel, Jeffrey, Carlos Vegh, and Guillermo Vuletin, 2013, On graduation from fiscal procyclicality, Journal of Development Economics, 100 (1), 32–47.
Freedman, Charles, Michael Kumhof, Douglas Laxton, Dirk Muir, and Susanna Mursula, 2010, Global effects of fiscal stimulus during the crisis, Journal of Monetary Economics, 57, 506–526.
Friedman, Benjamin, 1978, Crowding Out or Crowding In? Economic Consequences of Financing Government Deficits, Brookings Papers on Economic Activity, No. 3.
Gallatin, Albert, 1807, Report on the Finances November, 1807, (Annual Reports of the Secretary of the Treasury of the United States).
Gavin, Michael, and Roberto Perotti, 1997, Fiscal policy in Latin America, NBER Macroeconomics Annual, 12, 11–61.
Geerolf, François, 2017, Reassessing dynamic efficiency, manuscript, UCLA.
Ghosh, Atish R., 1995a, Intertemporal tax-smoothing and the government budget surplus: Canada and the United States, Journal of Money, Credit and Banking, 27, 1033–1045.
Ghosh, Atish R., 1995b, International capital mobility amongst the major industrialised countries: too little or too much?, The Economic Journal, 107–128.
Ghosh, Atish R., and Jonathan D. Ostry, 1997, Macroeconomic uncertainty, precautionary saving, and the current account, Journal of Monetary Economics, 40, 121–139.
Ghosh, Atish R., Jun I. Kim, Enrique G. Mendoza, Jonathan D. Ostry and Mahvash S. Qureshi, 2013, Fiscal fatigue, fiscal space and debt sustainability in advanced economies, The Economic Journal, 123, F4–F30.
Gorton, Gary B, and Guillermo Ordoñez, 2013, The Supply and Demand for Safe Assets. NBER Working Paper, No. 18732, National Bureau of Economic Research.
Gourinchas, Pierre-Olivier, and Olivier Jeanne, 2012, Global Safe Assets. BIS Working Paper, No. 399.
Grilli, Vittorio, Donato Masciandaro and Guido Tabellini, 1991, Political and Monetary Institutions and Public Financial Policies in the Industrial Countries, Economic Policy, 6, 342–392.
Hall, George J., and Thomas J. Sargent, 2014, Fiscal discriminations in three wars, Journal of Monetary Economics, 61, 148–166.
Hallerberg, Mark, Rolf Strauch, and Jurgen von Hagen, 2009, Fiscal Governance: Evidence from Europe, Cambridge: Cambridge University Press.
Hatchondo, Juan Carlos, Leonardo Martinez, and Francisco Roch, 2017, Fiscal Rules and the Sovereign Default Premium, manuscript.
Hausmann, Ricardo and Ugo Panizza, 2011, Redemption or Abstinence? Original Sin, Currency Mismatches and Counter Cyclical Policies in the New Millennium, Journal of Globalization and Development, 2, 1–35.
Heinemann, Friedrich, Marc-Daniel Moessinger, and Mustafa Yeter, 2018, Do fiscal rules constrain fiscal policy? A meta-regression-analysis, European Journal of Political Economy, 51, 69–92.
Huang, Yi, Ugo Panizza, and Marco Pagano, 2017, Local crowding out in China, EIEF Working Paper, No. 1707.
Huang, Yi, Ugo Panizza, and Richard Varghese, 2018, Does public debt crowd out corporate investment? International evidence, CEPR Discussion Paper, No. 12931.
International Monetary Fund, 2008, Fiscal Policy as a Countercyclical Tool., World Economic Outlook Chapter 5.
International Monetary Fund, 2012, Safe Assets: Financial System Cornerstone?, Global Financial Stability Report, Chapter 3.
International Monetary Fund, 2015, From Banking to Sovereign Stress: Implications for Public Debt, IMF Policy Paper.
International Monetary Fund, 2018a, Fiscal Monitor: Capitalizing on Good Times, April.
International Monetary Fund, 2018b, The Federal Democratic Republic of Ethiopia, IMF Country Report, No. 18/354.
Jackson, Matthew and Leeat Yariv, 2015, Collective Dynamic Choice: The Necessity of Time Inconsistency, American Economic Journal: Microeconomics, 7, 159–178.
Jordà, Òscar, and Alan M. Taylor, 2016, The time for austerity: estimating the average treatment effect of fiscal policy, The Economic Journal, 126, 219–255.
Jordà, Òscar, Moritz Schularick, and Alan M. Taylor, 2011, Financial crises, credit booms, and external imbalances: 140 years of lessons, IMF Economic Review, 59, 340–378.
Judd, Kenneth L., 1999, Optimal taxation and spending in general competitive growth models, Journal of Public Economics, 71, 1–26.
Kaminsky, Graciela, Carmen Reinhart, and Carlos Végh, 2004, When it rains, it pours: Procyclical macropolicies and capital flows, NBER Macroeconomics Annual, 2004, 11–53.
Kingston, Geoffrey, 1991, Should marginal tax rates be equalized through time?, The Quarterly Journal of Economics, 106, 911–924.
Kourtellos, Andros, Thanasis Stengos, and Chih Ming Tan, 2013, The effect of public debt on growth in multiple regimes, Journal of Macroeconomics, 38, 35–43.
Krogstrup, Signe, and Charles Wyplosz, 2010, A common pool theory of supranational deficit ceilings, European Economic Review, 54 (2), 269–278.
Krugman, Paul, 1988, Financing vs. Forgiving a Debt Overhang, Journal of Development Economics, 29 (3), 253–268.
Laeven, Luc, and Fabian Valencia, 2013, Systemic Banking Crises Database, IMF Economic Review, 61(2): 225–270.
Laubach, Thomas, 2009, New evidence on the interest rate effects of budget deficits and debt, Journal of the European Economic Association, 7 (4), 858–885.
Lijphart, Arend, 1994, Electoral Systems and Party Systems, Oxford: Oxford University Press.
Lizzeri, Alessandro 1999, Budget Deficits and Redistributive Politics, The Review of Economic Studies, 66, 909–928.
Lucas, Robert E., and Nancy L. Stokey, 1983, Optimal fiscal and monetary policy in an economy without capital, Journal of monetary Economics, 12, 55–93.
Marcet, Albert, and Andrew Scott, 2009, Debt and deficit fluctuations and the structure of bond markets, Journal of Economic Theory, 144, 473–501.
Mauro, Paolo, 2011, Chipping Away at Public Debt – Sources of Failure and Keys to Success in Fiscal Adjustment, London: J Wiley & Sons, Inc.
Mauro, Paolo, Rafael Romeu, Ariel Binder, and Asad Zaman, 2015, A Modern History of Fiscal Prudence and Profligacy, Journal of Monetary Economics, 76, 55–70.
Mauro, Paolo and Mauricio Villafuerte, 2013. Past Fiscal Adjustments: Lessons from Failures and Successes, IMF Economic Review, 61 (2), 379–404.
Mauro, Paolo and Jan Zilinsky, 2016, Reducing Government Debt Ratios in an Era of Low Growth, PIIE Policy Brief, No. 16–10, Peterson Institute for International Economics.
Mbaye, Samba, Marialuz Moreno Badia, and Kyungla Chae, 2018, Global Debt Database: Methodology and Sources, IMF Working Paper, No. 18/111.
Mc Morrow, Kieran, Werner Roeger, and Valerie Vandermeulen, 2017, Evaluating Medium Term Forecasting Methods and their Implications for EU Output Gap Calculations. Discussion Paper, Directorate General Economic and Financial Affairs (DG ECFIN), European Commission.
Mendoza, Enrique G., and Jonathan D. Ostry, 2008, International Evidence on fiscal solvency: Is fiscal policy “responsible”?, Journal of Monetary Economics, 55, 1081–1093.
Michie, Ranald, 2001, The London Stock Exchange: A History, Oxford: Oxford University Press.
Milesi-Ferretti, Gian Maria, 1997, Fiscal Rules and the Budget Process, Giornale degli Economisti, 56, 5–40.
Müller, Andreas, Kjetil Storesletten, and Fabrizio Zilibotti, 2016, The Political Color Of Fiscal Responsibility, Journal of the European Economic Association, 14, 252–302.
Neumeyer, Pablo A. and Fabrizio Perri, 2005, Business Cycles In Emerging Economies: The Role of Interest Rates, Journal of Monetary Economics, 52 (2), 345–380.
Olson, Mancur, 1965, The Logic of Collective Action: Public Goods and the Theory of Groups (Revised ed. 1971). Harvard University Press