Skip to content

Eurobonds, risk sharing and covid-19

[Continuing from the previous post discussing Coronavirus macro policy.]


The common nature of the coronavirus crisis has intensified calls for deepening European fiscal risk sharing. Whether that risk sharing takes the form of outright real time transfers, explicit bailouts in the future, debt-monetization through the Eurosystem or jointly issued eurobonds, is of secondary importance relative to the question whether this risk sharing itself is justified. Insurance is supposed to protect against idiosyncratic risk, and to the extent that an exogenous covid-19 shock might hit European countries with asymmetric strength, moral hazard concerns about “past irresponsible behaviour” are not justified. However, member states might also differ in their preparedness to a symmetric exogenous shock. Whether to share risks stemming from such asymmetric “pre-existing conditions” is less the domain of insurance, and more of redistribution, and solidarity. Looking deeper into the underlying reasons for these asymmetries helps us see why the moralizing narrative about past irresponsibility is overly simplifying, and why solidarity towards the euro periphery is warranted from northern members.


In an attempt to mitigate the economic cost of containment measures in their fight against coronavirus, European governments are set to spend massive amounts, leading to widening budget deficits and rising public debt. With the European Central Bank providing a monetary backstop to sovereign bond markets, no country should be constrained in being able to borrow the necessary amounts and having ample liquidity at their disposal. That said, while the ECB can keep interest expenditures low, the ensuing fiscal stimulus is still not a completely free lunch since the extra debt remains on the books of national governments and future taxpayers remain liable for them. In other words, liquidity provision (either through ECB bond purchases, or via ESM lending) does not equal transfers. Absent such transfers from outside, member states will need to increase the present value of their future primary budget surpluses by the amount of the extra debt issued now.


However, given that covid-19 is a common shock faced by the whole of Europe, and that there are substantial cross-border spillovers of one country’s containment measures (or the lack thereof), it is ever more widely suggested that Europe should pool their fiscal resources and respond jointly to this crisis. Such fiscal risk sharing would mean that even if the costs of containment are larger in one country than in others, these costs would still be shared among all member states and be paid for jointly since they serve the benefit of all. This would entail cross-border transfers (in net present value terms), as opposed to just lending and liquidity provision. This would take some of the burden off the shoulders of the taxpayers in countries which had to spend relatively more than other member states, meaning that their future fiscal policy will not have to be tightened as much as the crisis spending in their country would imply otherwise.


The form of these transfers is a somewhat independent question. They can be either

  1. outright real-time money transfers from less affected countries to more unfortunate ones; or 

  2. could take the form of a future bailout or debt restructuring whereby some of their national debt is written-off; or 

  3. whenever the risk of default threatens, the ECB (potentially via the ESM) can keep intervening to help roll-over maturing debt, in effect permanently monetizing it, constituting a “stealth bailout” through the balance sheet of the Eurosystem as opposed to an explicit debt write-off. 

  4. Perhaps best reflecting the spirit of this fiscal risk sharing would be financing the covid-19 spending with commonly issued eurobonds which member states are jointly liable for. The revenues raised from issuing such bonds would be spent where they are needed for the virus fight, but repayments would be shared proportionally across members, and the potential difference between these amounts would constitute cross-border transfers.


In the case of option D), eurobonds, opponents usually counter that with such form of debt mutualisation the transfers can potentially get larger than intended if a country defaults on their share of the repayments in the future – and that this default risk calls for fiscal union (where spending decisions are also jointly taken) before introducing a common debt instrument. While this argument has merits, if we work on the assumption that a member state on the verge of default will eventually be saved anyway via either methods A), B) or C), then method D) does not seem that much different. As for fiscal union, one can think of the fight against covid-19 as a limited form of “proto fiscal union”, where money raised from common debt issuance is spent only on the jointly agreed goals of mitigating the epidemic crisis. 


A more fundamental question than the precise form of this fiscal risk sharing, is whether such an insurance scheme is justified or not. Insurance is supposed to share idiosyncratic risk between parties. That is, if an unexpected event affects members in an asymmetric way, then more fortunate participants can help the unlucky ones, in exchange for expecting the same treatment were their roles reversed. Therefore, insurance agreements should be made “under the veil of ignorance” before each participant’s luck is revealed: ex ante insurance will lead to ex post transfers, but in expected value terms payouts should be zero. This is why “pre-existing conditions” are usually not insured, as this means the “veil is already lifted”. Another inherent issue with insurance is moral hazard, whereby the insured party engages in irresponsible behavior, making an accident and triggering an insurance payment more likely. Which is why insurance payouts are usually made conditional on not having engaged in such irresponsible behavior in the past. For example, building a house in the flooding area of a big river rather than on a hilltop could be considered as such irresponsible behavior. 


Moral hazard is the most widely used argument of those who oppose fiscal risk sharing in Europe, since according to them this would amount to reward irresponsible past overspending by profligate Southern member states. The proponents of risk sharing counter that covid-19 is an “exogenous shock” which is clearly not a result of anything what, for instance, Italy did but is rather an act of nature which justifies insurance. I would argue the truth is somewhere halfway between these two arguments.


Of course, if covid-19 ends up hitting Germany disproportionately less than an equally well prepared Italy (asymmetric exogenous shock), then insurance payouts to Italy are justified (even if we abstract from the earlier discussed positive spillovers of containment measures) since this can in no sensible way be affected by whatever Italy had done in the past. Before the crisis is over (i.e. before the veil of ignorance is lifted), it can go either way (e.g. Germany ending up as the recipient), so ex ante both parties should be willing to enter this risk sharing agreement as long as they are risk averse. It is similar to the case, when both of two homeowners responsibly built their houses on a hilltop, but only one of the hills were flooded by a huge torrent: the lukier one should help rebuild the other’s house. Instead, if they face a symmetric exogenous shock whereby both hills are flooded, and both houses get destroyed (both Germany and Italy are hit by covid-19 to the same extent), then it is basically an aggregate (as opposed to idiosyncratic) shock for which there cannot be insurance by construction. This is because both parties are equally affected and none of them can help the other. So no insurance payments will be made and everybody tries to rebuild on their own.


Now let’s assume it is a symmetric exogenous shock, but with asymmetric effects. The same flood engulfs both hills (i.e. covid-19 hits both Italy and Germany with the same ferocity), but one party had built their house on the lower segments of the hill and as a result got destroyed, while the other party had built on the topmost point and as a result survives. Italy entered the covid-19 crisis with very high public debt, therefore having less fiscal capacity to support the economy in the face of this exogenous shock, while Germany accumulated enough fiscal room in the previous years to be able to mount a massive stimulus package against the very same exogenous shock. In other words, the symmetry of the shock would require similar containment measures in both countries, so no insurance payments would need to occur between equally prepared parties. But Italy is less able to afford the same fiscal expansion than Germany is, not fully independently of its past. Italy has “pre-existing conditions” which make it less able to cope with the same crisis. Should these asymmetric effects of the otherwise symmetric shock, which are simply due to pre-existing conditions, be insured?


I believe, the proponents of risk sharing cannot answer this question just by pointing to the exogeneity of covid-19, and the fact that Italy’s past behaviour is in no way responsible for the epidemic. Of course, it is not the exogenous epidemic itself or its potentially asymmetric distribution for which Italy could be responsible, but rather for its preparedness to the same exogenous epidemic. Much like the unfortunate homeowner is not responsible for the exogenous flood itself, but rather for building their house on a more dangerous part of the hillside. In other words, the answer to the question should not emphasize the exogeneity of the shock, but instead investigate why the pre-existing conditions, which lead to asymmetric effects, have developed.


Asymmetries and pre-existing conditions which are known with certainty ex ante, technically cannot be insured, since that would violate the zero expected value requirement (i.e. the “veil is already lifted”). Cross-party payments can still be made, but those are better called redistribution instead of insurance. Deciding whether redistribution is justified is less straightforward than in the more objective risk-reward framework of insurance. It involves deciding also on the very normative notions of what we consider to be “fair”, and depends on how strong the perception of “belonging together” and solidarity is between the parties. We should try to ascertain whether pre-existing conditions are really due to “past irresponsible behaviour” or they are the just cumulative result of past uninsured acts of nature. It is usually difficult to give a very clear-cut answer to that, and it necessarily involves a great deal of judgment, but let’s see where it leads us.


Returning to the example of the hill people, if the homeowner with the destroyed house decided to build on the riskier lower hillside just in order to avoid the chilly hilltop temperatures and enjoy the hillside sun, then we’d think compensation payments would be less fair after his house is flooded. In contrast, assume he would really have wanted to responsibly build on the hilltop, but the other party happened to own the top property also on this hill, and did not allow building anything on it. Then we’d expect the other party to help with the costs of rebuilding. Similarly, even if the other party had nothing to do with the misfortune of the unlucky one, it can still be that there were differences in opportunity: assume the unlucky did build on the top of the hill, it’s just that his hill was not high enough. There is a pre-existing condition, but it is really not the result of past irresponsible behaviour, just of past luck, so we might still find it fair to help the less fortunate party.


In the case of Europe, if Italy’s currently weak fiscal situation is the result of them “irresponsibly living beyond their means” in the past, then it’s easier to refuse ex post insuring this profligacy. But what if the situation is a bit more complex than that? I have written about this in more detail before, but here are the highlights. 


What if today’s Italians really did all which can reasonably be expected given the huge debt pile they inherited from their grandparents? Italy has been running primary budget surpluses for the past three decades: this is hard to square with the image of a spendthrift country incapable of fiscal restraint, which Italy is painted to be.


What if the lacklustre post-crisis recovery in the periphery (and the correspondingly high debt-to-GDP ratios) are not just the result of laziness and lack of structural reforms, but also of excessive austerity imposed on them by overly strict creditors such as Germany, and by the rigid fiscal rules of the eurozone? The periphery gave up their monetary sovereignty, constraining their own fiscal space (no own central bank to provide backstop any more) in the hope that this will be replaced by common collective help if the need comes. Would it be fair to deny this?


Another major contributor to recently slow growth in the periphery were the tight fiscal stance and weak demand elsewhere in the euro area, notably in Germany whose record big current account surpluses reflect excess savings and low domestic spending. In a liquidity trap, the irresponsible ones are not those who borrow, but those who do not spend (“virtue becomes vice, and prudence is folly”). During such times, CA-surpluses deprive the rest of the eurozone from aggregate demand and unleash recessionary forces on trading partners. It is hard to grow out of your debt and/or save more when your main trading partner is not spending enough so you don’t have enough income. Therefore, looking at Southern indebtedness in isolation of German excess savings makes little sense: they are two sides of the same coin.


Similarly, what if Germany’s ability to combine a high savings rate with low unemployment before the financial crisis was enabled by the willingness of the euro periphery to borrow and spend? Without sufficient external demand from trading partners, weak domestic spending in Germany would have led to higher unemployment, depriving them of the very incomes from which they can save. Blaming the source of their savings for profligacy, while priding themselves on how much they saved does not seem like a consistent position: since one’s borrowing is the other’s saving it is impossible for everyone to save at the same time. Also, what if the periphery’s credit bubble was fuelled by irresponsible lending from German banks, in search of high yielding assets to absorb domestic savings, as opposed to by excessive borrowing demand in the periphery? 


Alternatively, what if Germany’s ability to run large export surpluses and accumulate savings was enabled by a weak euro exchange rate, which in turn was too strong for the periphery? What if by giving up their own currency, the periphery, in effect, sacrificed their export competitiveness to support Germany becoming the “Exportweltmeister”? Or what if it’s not Southern countries’ unit labor costs which rose too much, but instead it’s Germany’s which has fallen too low? The inflation-adjusted unit labor cost (i.e. labor’s share of income) remained fairly stable in the periphery (meaning that real wages grew in line with productivity), while it steadily decreased in Germany, as German workers’ share of GDP was gradually depressed and redistributed to capital owners. Is rising inequality what a “responsible” South was supposed to engineer, too?


Considering the above aspects could nuance our view about euro periphery countries with pre-existing conditions “deserving” solidarity and inclusion in fiscal risk sharing for the fight against covid-19 – relative to the simplistic and moralizing narrative whereby big debts can only mean past irresponsibility. Whatever our answers are to these questions though, failing to display unity in the face of such a grave crisis would surely wound our European integration for a substantial time to come. A responsible country must not want this.

Published inArticle