.@M_C_Klein and @martinwolf_ don't understand what's wrong with Vollgeld and related proposals, because they aren't working with a good model of money. 
The claim is that the financial system "unnecessarily" and "by accident" links essential public services to private risk-taking. That's a strong claim. And it's wrong.
Modern economic growth was born with the modern financial system and the cradle was 1750 to 1875 Britain. This is simple fact.
Economists and historians have not tied the two together properly, because they haven't been working with the right models.
There's a basic problem of finance and of debt contracts that borrowers want loans that are long term and unsecured, while lenders want loans that are short term and grant the right to foreclose on assets at the earliest sign of default.
Most legal systems get the balance wrong, so debt is unproductively low, and there are a lot of gains that the economy will never realize.
Banking developed -- in a world where bankers had unlimited liability -- as a means of balancing the interests of borrowers and lenders and of massively increasing the productivity of the economy by massively increasing the quantity of *productive* debt in the economy.
Banks don't just transform maturity (i.e. offer short term liabilities to lenders and long term loans to borrowers), they also are (or with reference to the modern experience should be) incentivized to transform maturity successfully or lose money.
A basic understanding of a bank's capital structure tells you that it doesn't "intermediate" between borrowers and lenders, it stands between them with its equity and subordinate debt protecting borrowers from losses.
Losses to non-bank creditors of the early British banking system were tiny -- precisely because the losses all fell on the bankers themselves. There was good reason to trust the banks, and there was good reason for the debt-based economy to grow.
This was all possible because of extraordinary legal innovation, particularly in the law governing negotiable instruments, which still underlies our financial system today.
There was one problem with this system. Being a banker was risky. So the system only really took off after 1763 when the Bank of England was just beginning to understand what a lender of last resort was supposed to do.
In the event of a liquidity shock and temporarily tumbling asset values, the central bank could step in to support the banks and make it possible for them to "hide" the illiquid assets on their balance sheets until the shock was over.
It's important that all central bank loans were recourse loans, so they did not "bail out" banks. If the assets didn't recover from the temporary illiquidity situation, they were still posted as losses to banks. (There's a right way and a wrong way to do lender of last resort.)
It was only when bankers began to trust in the Bank of England to give them this support that banks really began to pop up in big numbers, that productive debt became common and that the frictionless neoclassical model (Adam Smith) could be imagined.
So why do we have banks? Because we need someone who can go bankrupt to manage the very delicate system of maturity transformation. A debt-based economy functions because the bankers are the ones who lose if they mess up.
And why do we have a central bank? Because that's what protects and stabilizes the asset prices imagined in the neo-classical model. It does so by helping banks hide illiquid assets *temporarily* on their balance sheets.
When you shift to "market-based" lending and do everything you can to reduce the banking system's support of the financial system, you create a world where we aren't protected from liquidity shocks and where asset prices are much less stable.
Bank runs were never a sign of a problem with banking itself. The are just a sign that banking is the means we use to deal with the fundamental liquidity problem that is inherent to economic activity.
Arrgh! I posted these tweets accidentally before I proofed them. That should be protecting *depositors* from losses.
In order to understand what banks do, you have to ignore the neo-classical model and work with a model that assumes that liquidity frictions are ubiquitous. #NewMonetarism
