Let’s say that somehow — miraculously — the United States achieved a balanced budget, and furthermore assume that this did not interfere with cyclical goals. There would be many fewer T-bills, especially since the current structure of the debt is quite short-term.
Fully safe collateral would be hard to come by.
I see the paucity of safe assets, and safe collateral, as a major financial problem looking forward. Our economy desires to extend more short-term credit than we can back by ready safe collateral or that we can cover by FDIC-like insurance. Yet credit moves forward, in part because of bailout incentives but also because failed managers simply aren’t, and cannot be, punished very hard. We return to the 19th century problem of bank runs, this time on the shadow banking system, and we realize to our horror that 1934-19?? was the exceptional and now-disappeared “safe period.” We shudder to think of the next crisis, which is one reason why so many people are skittish about the Greek situation.
T-bills limit one agency problem, but create another. If there were T-bills “coming out our noses,” finance would be much safer, although of course we would have to face the moral hazard problem of what the government does with so much borrowed money. When the quantity of T-bills goes down, financial regulation becomes a tougher act to pull off.
When the Fed pays interest on reserves, cash becomes like T-bills. Interbank lending falls, because there is less need to dispense of “idle” reserves. I agree with Scott Sumner that interest on reserves is contractionary and that has had negative macroeconomic consequences. Still, I worry that if we eliminate interest on reserves, the regulation of interbank credit becomes more problematic. Much more overnight lending would go on. In essence fear of inadequate bank regulation pushed the Fed to contract in this manner.
If you want to get rid of all or nearly T-bills, you are very optimistic about bank regulation. I am not very optimistic about bank regulation.