Economics professor Miles Kimball has a new blog, Confessions of a Supply-Side Liberal. In one of his first posts, he outlines a plan for stimulus that he calls “Federal Lines of Credit” (FLOC). It's presented in a longer policy paper, “Getting the Biggest Bang for the Buck in Fiscal Policy." This has gotten interest across the political spectrum. Bill Greider has written about it in The Nation, as has Reihan Salam in the National Review.

What's the idea? Under normal fiscal stimulus policy in a recession, we often send people checks so that they'll spend money and boost aggregate demand. Let's say we are going to, as a result of this current recession, send everyone $200. Kimball writes, "What if instead of giving each taxpayer a $200 tax rebate, each taxpayer is mailed a government-issued credit card with a $2,000 line of credit?" What's the advantage here, especially over, say, giving people $2,000? "[B]ecause taxpayers have to pay back whatever they borrow in their monthly withholding taxes, the cost to the government in the end—and therefore the ultimate addition to the national debt—should be smaller. Since the main thing holding back the size of fiscal stimulus in our current situation has been concerns about adding to the national debt, getting more stimulus per dollar added to the national debt is getting more bang for the buck."

Let's kick the tires of this policy. There's a lot to like about the proposal, particularly how it could be used after a recession is over to provide high-quality government services to the under-banked or those who find financial services yet another way in which it is expensive to be poor (modified, it turns right into Steve Waldman's Treasury Express idea). It's not clear whether this is meant to supplement or replace normal demand-based fiscal policy - at one point he proposes it could balance out a "relatively-quickly-phased-in austerity program."

As a supplement it has promise, but I think there are some major problems with this proposal, which can be grouped under four categories.

I: Isn't deleveraging the issue? Is this a solution looking for a problem? From the policy description, you'd think that a big is credit access holding the economy in check.

But taking a look at the latest Federal Reserve credit market growth by sector, you can see that credit demand has collapsed in this recession. Consumer credit drops throughout the beginning of the recession, particularly in 2009. This is true even for consumer credit by itself, which rebounds in 2011. It's not clear that these lines of credit would be used to expand demand at the macro-level; likely, given what we see, it would be used to replace other, higher-interest forms of debt (see III), a giant transfer of credit risk from credit card companies to taxpayers. But certainly some people will benefit, so let's examine why this policy is supposed to work.

II: This policy is like giving a Rorschach test to a vigilante. No, not that vigilante. I mean the bond vigilantes. Because to assume this plan would work, you need to make some curious assumptions about how bond vigilantes think, as it increases the debt by a significant amount.

Let's say our country has a balanced budget with a debt-to-GDP ratio of 50 percent and we hit a recession while at the zero-bound. As a result of less tax revenue coming in and more automatic stabilizers going out, debt-to-GDP will be 60 percent at the end of the year. We want to stimulate the economy further using fiscal stimulus.

Let's say our default is that we take three percent of GDP, divide it among the population, and mail it out. At the end of the year, the debt-to-GDP ratio will be 63 percent (I am ignoring that fiscal stimulus at the zero-bound can be largely self-financing for this example).

In Kimball's FLOC, we instead take 9 percent of GDP, divide it evenly among the population, and mail out lines of credit that add up to that 9 percent of GDP. Let's also say that perfect forecasting tells us that within the year, 6 percent of it will be utilized as a loan not yet paid back, and 3 percent is still available as credit.

What's the government's debt-to-GDP ratio at the end of the year in Kimball's example? I'm not sure how he'd account for it. I imagine it should be 69 percent (60 + 9). Perhaps it is 66 percent (60 + 6)? Either way, it is more than the 63 percent of just giving people money. His plan requires a larger debt-to-GDP ratio. If his accounting ends up with just 60 percent, I'm not sure I understand how he is doing it.

Now Kimball will say that bond vigilantes will be happy with this. Why? Because there's a built-in plan for repaying it. "[T]he fact that much of the money would ultimately be repaid would dramatically reduce the ultimate addition to the national debt...(though at a relatively attractive ratio of additional aggregate demand to addition to national debt)."

If we are guessing as to what the bond vigilantes want, it is clear they want more U.S. government debt. Ten-year Treasuries are selling at 1.5 percent, while real interest rates are negative! But for the purposes of the FOLC, we need a few assumptions about what the bond vigilantes think, which aren't clear.

First (i) it assumes that the bond market will only care about the government's long-run debt ratio instead of the short-term. I think that's correct. But much of the bond vigilante argument is predicated on the opposite, that no matter what the long-term is, the capital markets will freak on short-term deficits.

It also assumes (ii) that the repayments of these FOLC will be made easier through debt collection than just collecting the equivalent amount of money through taxation. I see no reason why that's the case, and many reasons to believe the opposite.

III: This policy will involve trying to get blood from a turnip. I very much distrust it when economists waive away bankruptcy protection. Especially for experimental, controversial debts that have never been tried in known human history.

As the paper admits, this is a machine for generating adverse selection, as the people most likely to use it are people whose credit access is cut due to the recession. High-risk users will likely transfer their balances from higher rate credit cards to their FOLC (either explicitly or implicitly over time if barred) - transferring a nice chunk of credit risk from the financial industry to taxpayers.

It's also not clear what happens a few years later when consumers start to pay off the FOLC. Could that trigger another recession, especially if the creditor (the United States) doesn't increase spending to compensate?

The issue isn't whether or not the government will be able to collect these debts at some point. It has a long time-horizon, the ability to jail debtors and use bail to pay debts, the ability to seize income, old-age pensions and a wide variety of income, and the more general ability to deploy its monopoly on violence. The question is whether this will be smoother, easier, and more predictable than just collecting the money in taxes. We have a really smooth system for collecting taxes, one at least as good as whatever debt collection agencies are out there. If that is the case, there's no reason to believe that this will satisfy the bond vigilantes or bring down our debt-to-GDP ratio in a more satisfactory way.

IV: Since we've very quickly gotten to the idea that we'll need to jettison legal protections under bankruptcy for this plan to work, it is important to emphasize that this policy is the opposite of social insurance.

I don't see a macroeconomic difference between the government borrowing 3 percent of GDP and giving it away and collecting it through taxes later versus the government borrowing 3 percent of GDP, loaning it to individuals, and collecting it later through debt collectors except in the efficiency and the distribution.

The distributional consequences of this proposal aren't addressed, but they are quite radical. Normally taxes in this country are progressive. Some people call for a flat tax. This proposal would be the equivalent of the most regressive taxation, a head tax. And it also undermines the whole idea of social insurance.

Let's assume the poorest would be the people most likely to use this to boost or maintain their spending. I think that's largely fair - certainly the top 10 percent are less likely to use this (they'll prefer to use high-end credit cards that give them money back). This means that as the bottom 50 percent of Americans borrow and pay it off themselves, they would bear all the burden for macroeconomic stability through fiscal policy. Given that the top 1 percent captured 93 percent of the income growth in the first year of this recovery, that's a pretty major transfer of wealth. One nice thing about tax policy, especially progressive tax policy, is that those who benefit the most from the economy provide more of the resources. This would be the opposite of that, especially in the context of a ""relatively-quickly-phased-in austerity program."

Efficiency is also relevant - as the economy grows, the debt-to-GDP ratio declines, making the debt easier to bear. The most likely borrowers under FOLC, the bottom 50 percent, have seen stagnant or declining wages overall, especially in recessions. A growing economy would keep their wages from falling in the medium term, but this is still a problematic issue - their income is not more likely to grow to balance out the payment burdens than if we did this at a national level, like normal tax policy.

The policy also ignores social insurance's role in macroeconomic stability, and that's insurance against low incomes. Making sure incomes don't fall below a certain threshold when times are tough makes good macroeconomic sense and also happens to be quite humane. This is not that. As friend-of-the-blog JW Mason said, when discussing this proposal, the FOLC is like "if your fire insurance simply consisted of a right to borrow money to rebuild your house if it burned down."

What else am I missing about this proposal?