Between the euro crisis, the refugee crisis, tensions within the single market, and anti-European political extremism, the European Union appears on the surface to be failing. This isn’t quite true though. Every time Europe faces a setback, it tends to make progress in response. This progress is usually only partial (or incomplete), but it is enough to lay the foundations for more comprehensive solutions to emerge in the future. What looks like failure is actually ‘failing forward’, a dynamic that Dan Kelemen, Sophie Meunier and I examined in a recent article in Comparative Political Studies. The latest incarnation of this concept is the recent developments – or lack thereof – at the December 2015 European Council summit, which was supposed to shore up European financial markets by pushing ahead with the construction of common institutions to safeguard European banks.

Judging from its pronouncements, the last European Council summit seems to have actually moved backwards on achieving a banking union. Although the Presidency Conclusions call for ‘work to advance rapidly’, they fail to mention the crucial role of pooled deposit insurance. This is in contrast to the 7 December draft of the Presidency conclusions, which called for ‘the gradual introduction of a European Deposit Insurance Scheme’ – presumably in line with the plan announced by the European Commission on 24 November. Deposit insurance is vital to the completion of banking union because it prevents customers from pulling their money out of banks whenever there is a financial crisis – like what happened to the Northern Rock in the United Kingdom in 2007, the Icelandic banking system in 2008, the banks of Cyprus in 2013, and the Greek banks last spring and summer.

The reason deposit insurance was omitted from the Presidency Conclusions is no secret. The German government is opposed to any cross-border redistribution of banking risk. So are the governments of many other Northern European countries. This opposition has been manifest since the banking union proposal emerged in early 2012. Strong assertions by economists at the European Commission and elsewhere that pooled deposit insurance is a necessary condition for stable financial market integration have had little influence on the conversation.

The fact that the European Commission announced an ambitious proposal for pooling deposit insurance is arguably more surprising than the fact that the European Council ignored it. Whatever the technical merits of pooled deposit insurance, national interests and preferences determined what made it into the summit negotiations. Indeed, reference to deposit insurance disappeared from the draft Presidency Conclusions on 14 December even before the summit took place.

But that is not going to be the end of the story. To see why, you just have to look at what is happening in Italy. Last month, the Italian government finally took steps to resolve four small banks with weak balance sheets. To give a sense of the magnitudes here, the four banks account for less than one percent of total banking assets in Italy. The problem, though, is that the balance sheets of these banks were very weak. As a result, it would be impossible to resolve the banks and to protect depositors without calling upon additional resources.

The choice faced by Matteo Renzi’s government was whether to get the financial resources necessary to shore up the four banks from tax payers via a state bailout or to get them from other banks. The Renzi government chose to get the resources from other banks. This is what the European institutions and other European governments insisted upon as well. And getting resources to shore up failing financial institutions from other banks is essentially what pooled deposit insurance is all about.

The question is how big or deep the pool should be. You can answer this question politically or you can answer it economically. The political answer is to say that the pool should be defined by sectoral, regional, or national boundaries. This is essentially the German position and it is a good answer insofar as you can make a political argument to explain why banks of a particular type or from a particular place should support one-another. But it is a bad answer insofar as it offers no guarantee that the banks within a particular sector or from a particular place will actually have the resources to bail out or wind up the institutions that get into trouble. That is the lesson we learned in the United States during the savings and loan crisis; it is the same lesson that we learned in Europe through the experience of countries as diverse as Iceland, Cyprus and Spain.

The economic answer to how deep the insurance pool for banking deposits should be is to look at the size of the market rather than at the type or location of the banks involved. If we are trying to shore up banks that compete across the European Union, then we will need to pool resources across the EU to come up with an adequate level of support. Even then, the insurance could prove inadequate and so policymakers are likely to need a backstop. This is what we learned from the 2012 Spanish banking crisis; it is also the reason the conversation about banking union started in earnest.

Hence the challenge is to convince sceptical European policymakers (and their even more sceptical electorates) of the true scale of the economic problems associated with deposit insurance. This is a multi-step process. The first step is to highlight the problem. Then it is necessary to outline a solution, to show the scale of the effort involved, and to provide evidence that is clear and compelling enough for politicians to be able to overcome any political opposition. Policy change within this model is a matter of trial and error – where making mistakes reveals both how things can be done better and the cost of doing nothing at all.

The European institutions are relatively far along in this process of shaping and justifying a banking union. Ultimately, that banking union will include some form of deposit insurance. It probably won’t be elegant and it will always need a backstop, but Europe will have to build some kind of bulwark to prevent bank runs and to stop bank bailouts from bankrupting national governments. If that deposit insurance does not happen before the next crisis, then politicians will just have to learn the hard way that pooling resources across countries is better than the alternative. That is a disappointing message for all of us who are condemned to live in the present and hence condemned also to experience the next crisis first hand. But it is a positive message for those who look to understand patterns (and progress) over time.

European policy-makers may be ‘failing’ at the moment – if by failure we mean refusing to adopt what economists describe as optimal policies – but they are making progress insofar as they now recognize the problem and are coming to grips with the implications of designing and implementing an adequate solution. Half measures introduced at any given summit often fail to solve the problems Europe faces. Just as often, though, those same failing institutions lay the foundations for more comprehensive solutions to emerge in the future.

The problem is that Europe’s policy-makers are having difficulty bringing the voters along with the wider European project. They are particularly bad at explaining both the goal of integration and the method. Voters tend to focus too closely on the present as a result – particularly when they perceive a threat to their life’s savings or watch helpless migrants risk death by drowning.

What Dan, Sophie and I describe as ‘failing forward’ looks a lot like muddling through on a good day. On a bad day, it just looks like failure. As academics, we can show that this is the way European  makes progress. Alas, for politicians looking to shore up support for European integration that not enough. ‘Failing forward’ is a hard message to sell to a public seeking reassurance from their leaders. Policy-makers will have to come up with something better to justify their actions.

Follow @Erik_Jones_SAIS