The Greek referendum has left the Governing Council of the European Central Bank (ECB) with a political choice that it should not have to make. The ECB will need cover from Europe’s political leaders no matter how this plays out. As with most important choices, this one will make some people very unhappy. We should expect to see opposition emerge both in the media and in the courts. Worse, the choice that the ECB has to make will unfold in stages. It involves a series of decisions and not a simple one-off commitment. That means Europe’s political leaders will have to insulate its central bankers from opposition for the foreseeable future and probably until long after the immediate crisis has passed. Finally, this is a choice that will define Europe’s future; not only will it tell us precisely what it means to be a member of the euro as a single currency, but it will also set a precedent for how much solidarity national governments should expect to receive and to offer.

At the very core, the choice that the ECB has to make is whether or not to provide the Greek banks with enough liquidity to honour their deposits. That liquidity will most likely take the form of emergency liquidity assistance from the Greek central bank. This is not a simple decision. The ECB has to offer enough emergency liquidity assistance to reassure Greek depositors that they will be able to get access to their money as and when they need it. That is not possible under the cap that the ECB put in place once the Greek negotiations broke down. That is why Greek deposit holders are limited to withdrawing just €60 per day. When the banks finally reopen, they will want to be able to take out more. Only the ECB can make that possible. It will have to decide whether or not to do so.

Again, this is not a one-off decision. If the ECB decides to increase the liquidity that the Greek banks can access, it also has to decide to accept the collateral that they have to offer. This is challenging because most of that collateral is backed by the Greek government. The Greek government missed a payment to the IMF on 30 June and it is no longer within a European bailout program. The referendum outcome does not change either of those conditions. Hence it is an open question whether the ECB should continue to accept Greek government-backed assets as collateral. The ECB could delay giving an answer while the referendum was on the horizon; now that the referendum has passed, the ECB will need some new justification for pretending that the events of 30 June did not happen or do not matter. That is why the Governing Council chose to increase the discount (or ‘haircut’) it requires to accept Greek-government back collateral the day after the referendum took place. That decision was just an acknowledgment of reality and not a clear sign of what is to follow. Moreover, the Greek government has a number of other payments to make for which it does not have the resources – including the repayment of bonds held by the ECB itself on 20 July. If the Greek government were to miss that repayment, then the ECB would have to justify how the Greek banks could use government-backed assets as collateral given that the Greek government had already failed to honour its obligations. Yet without using those assets, the Greek banks would have insufficient access to liquidity no matter where the ECB sets the cap on its emergency assistance.

If the ECB decides to cut off the liquidity to the Greek banks – or to continue restricting it, which would amount to the same thing given how much liquidity the Greek banks require to re-open – then the question the ECB faces is how to minimize the threat of contagion to other countries in southern Europe. The conventional wisdom in the markets right now is that the ECB has the equipment in its arsenal, including both the promise to buy unlimited amounts of bonds from countries that get into trouble and the broader policy of buying assets from across Europe as part of its program for quantitative easing. This conventional wisdom assumes that these instruments work automatically. Unfortunately, they do not.

The ECB can only buy unlimited amounts of debt from countries that get into trouble under specific conditions. The governments of those countries have to ask for help, they have to agree to accept supervision over their economic policies, and they have to retain access to private capital markets. Unfortunately, these are not easy conditions to meet. Most governments do not want to admit to the markets that they have a problem – which is what asking for help from the ECB does. Even if they are willing to admit that they have a problem, they do not want to accept European supervision over their economic policymaking. And they have little time to negotiate over terms and conditions because they face the risk that they will lose access to private capital markets. In other words, what sounds like a powerful instrument is actually very difficult to use in practice. That is why the Spanish government refused to ask for help from the ECB in the autumn of 2012 – even though solving the Spanish crisis was precisely why the ECB created this instrument in the first place.

If the ECB cannot use this more targeted policy of propping up individual governments, then it will have to rely on the more general policy of quantitative easing. There too, however, it is likely to run into trouble. Quantitative easing was only embraced by the Governing Council on two conditions. First, the purchases should be roughly proportional to the size of the economies in the euro area – in other words, if the German economy is roughly 1.5 times the Italian economy, then the ECB should buy €3 of German assets for every €2 of Italian assets. (Actually, the German economy is even larger relatively to Italy, but the numbers are easier this way). This is a problem because when investors get scared in Europe, they are likely to sell Italian assets and buy German assets. This means not only that the ECB will have to add to demand for German assets at a time when the price is already rising but also that it will not be able to add enough demand for Italian assets to close the spread between Italy and Germany. On the contrary, the spread will widen unless the ECB chooses to ignore its obligation to buy bonds proportionately – which it can do, but only with adequate political cover.

Even if the ECB chooses to buy more Italian assets than German assets, it will become relevant to ask who will cover the losses if the policy fails to calm the markets. This is where the other provision written into the ECB policy of quantitative easing becomes important. The risk associated with purchases of Italian bonds should be held within Italy. That means that the more effort the ECB has to put into buying Italian bonds to calm the markets, the more risk the Bank of Italy must assume for a potential default by the Italian state. You don’t have to believe that the Italian state will actually default to see that this is a worrying dilemma for market participants. Beyond a certain threshold – and we don’t know where that is precisely – the increase in the potential losses from Italian government bonds held by the Bank of Italy will become more worrisome than the purchases of additional Italian bonds will be reassuring. Hence the ECB will have to find some way around this national responsibility for assuming the risk from national bonds to make sure that the crisis in Greece does not spread.

However this turns out, the ECB will need to make a choice for Europe to avoid catastrophe. It can choose to save the Greek banks and hold Greece in the euro; it can choose to bailout countries that get into distress without forcing them into humiliating admissions of weakness or conditions for support; it can act more stealthily to purchase assets without strict proportionality; or it can finally admit that the risks associated with ECB activities belong to all euro area countries and not just those that issue the assets the ECB purchases. These are all political choices and so arguably go beyond the mandate of the ECB. But they are necessary choices and so Europe’s politicians must be prepared to back whatever the ECB thinks is most sustainable. What we know now from harsh experience is that making no choice is actually the worst choice of all.

This essay is published in Italian by Il Foglio.

Follow @Erik_Jones_SAIS