This post aggregates some comments I made in previous posts. Thanks to questions and comments by the likes of Arthurian, Mario, and Hans, I'm able to to further refine my points, and try to understand better even my own specific views and position.
Money from government a.k.a. government deficit spending is added money that was not borrowed into existence. Whether it goes through a bank (via funding a bank loan), or government pays you direct (doubtful), it adds to reserves/deposits. Creating/printing money via a bank loan is riskier, and eventually gets unprinted when it's paid back. 'Printing' via government spending never has to be paid back if it was paid for services rendered, and therefore, more permanent.
Writedown of upside down private sector debt
Printing money to extinguish private sector debt seems like a good plan at first, but it doesn't take into consideration that a lot of the debt was incurred to finance non-productive endeavors, such as asset speculation. In other words, this plan would be like printing money without increasing productive capacity. This would likely lead to inflation. It's so much better to print money in exchange for work which leads to productive capacity. Plus, paying people for their debts is unfair. What would that tell people going forward? That it pays to get into debt? And what if people in the future get excessively into debt in order to get ahead, i.e, bid up asset prices while trying to corner assets in the economy. In the end, you'll have people with all the assets and no debt, and then people who never went into debt and don't have any assets. Not fair, and could lead to gross misallocation and crises.
But then, I see that this scenario causes the newly-printed money to be destroyed (along without the debt), and hence not cause inflation. It also frees private sector from the peachy burden of deleveraging, and get private sector demand moving again. Therefore, I gee that there should be an element of debt writedown in solving the current mess, but there should also be a rule that says people forgiven their debt should not turn around, and use their now increased net equity to again speculate on more assets. after all, with their debt erased, they can now sell their assets and bid up on new assets,(and especially if NGDP proponents are listened to, and money is thrown to people to BORROW again. After all, if people don't borrow, how does NGDP grow constantly?) If this scenario happens, we'll now have inflation, and we're right back to people with heavy debt.
End fed distortion of rates
I'm not sure if the Fed should do anything anymore to move rates, even if to make it go up. Right now, I'm thinking just keep rates at zero, but neither should it pay interest on reserves. This way, private sector banks don't make lending decisions based on fed rates, but on real market considerations. This is hardly a set position for me yet. I'm not sure if there's a downside to this approach, but I think Warren Mosler is proposing something on this line, and I think it considers a lot of the unseen pitfalls and distortions that interest rate setting does to the financial market.
The Fed just could lend at zero to a bank that needs reserves for payment purposes. It could also lend at zero to a bank that wants to lend, but it will now need to insist on a high capital equity ratio, higher than what it is now (Capital ratio pertains to bank capital. This is owner's equity that will bear the first risk of loss if the bank ends up with a soured loan, before depositors start losing their money. Basel recommends a loan ratio of no more than 15 times capital. Bigger than that, the depositors' money starts getting more at risk).
Increase bank capital requirements
Banks should hold a larger chunk of capital as buffer agains loan loss, and that should be the first to get a haircut when loans go sour. It will then be the cost raising equity capital that will keep the banks from making loans that could just be misallocated. The pain of losing their own equity therefore should be strong enough to keep them from making sketchy loans, despite zero fed rates. If any bank is left with any excess reserves, they shouldn't get any wild ideas, by lending it to the first subprime who promises high yield.
Fed setting of rates is too distortionary, and may result in misallocation. Having a mandated loan capital ratio of between 5 to 10 times capital helps keep excessive risk-taking at bay, without creating boom busts associated with gaming the interest carry spread.
The bank owner should have main responsibility for solvency and prudence, that's why there should be enough owner's capital to begin with. However, with banks now owned hundreds of thousands of nameless shareholders, while bank decisions are made by executives whose only skin in the game is their bonus, this ideal is easier said than done. And yes, the government ends up having the main responsibility, when banks with insufficient capital take on more risk than they can. But I don't think keeping banks private (as opposed to widely-held) is a good contravening measure, even though being a public company dilutes bank ownership (and dilutes the prudential control and prudence that comes with concentrated ownership). If you prohibit banks from going public, it becomes harder for them to raise capital, and hence, to grow loans to help grow the economy during expansions (the more equity, the more loans they're allowed to make).
Better regulation and risk monitoring
The problem of widely-held banks should be solved by better regulation of executives. You can't run banks the way you run companies that maximize revenues. Executives who think sales are the prime objective are just talking their book, and thinking of their bonus. Because of the implicit government commitment to protect it, it should be run like a utility, and with risk management as prime objective.
The fed has a bias to protect banks because that was its original mandate, to be the lender of last resort. In an ideal world, the Fed monitors (and understands) the kind if risks banks take, because they will have to come in and rescue them if a run happens. Though Fed-injected reserves help against a bank run, it is a loan from the fed to the bank. It has to be paid back by either raising capital or getting back the depositors. If a bank is unable to, the fed essentially bails out the bank.
I'm not there yet on taking off the cap on FDIC. I think this will increase moral hazard among banks. Banks may increase their rates to depositors, knowing people would put their deposits on any bank that promises the highest rates, because they know all their money is guaranteed by the government. Although as mentioned, I'm not sure yet how the dynamics would change if rates are permanently at zero. 100% FDIC guarantee may induce banks to compete on higher rates to depositors, but then again maybe not.
...more discussion in the comments...