Consider the following thought experiment. The government imposes a tax of $1000 on all bankers. One the very same day, the government authorizes a new spending program, a $1000 subsidy to all bankers. How should we think of this combined policy? To me, it’s a nothingburger.
Economists used to view reserve requirements as an implicit tax on banks. That’s because in the old days there was no interest paid on bank reserves, so there was a high opportunity cost of holding reserves.
Now, we have no reserve requirements, but we do pay interest on reserves (IOR). This was done because policymakers wished to move to a “floor system”, where banks would choose to hold large quantities of reserves. The adoption of IOR allows the central bank to inject lots of reserves into the system, without driving interest rates down to zero. You can think of large reserve holdings as a tax on banking, and IOR as an offsetting subsidy.
Chris Giles has an article in the FT where he suggests that the BoE move to a system where the tax is maintained but the subsidy is removed:
The central bank pays 5.25 per cent on reserves so that it can set the short-term policy interest rate at that level. It is effective, but not the only way to control short-term rates.
Instead, it could require banks to hold a fixed amount of money without interest, paying 5.25 per cent only on a small part of the reserves.
I don’t like the idea of paying interest on reserves, but I also oppose reserve requirements.
Go back to the thought experiment at the top of this post. Suppose the government suddenly removed the $1000 subsidy to bankers, but kept the $1000 tax in place. How should we think about that change? In a technical sense, it involves a cut in government spending. But it also moves us from a situation where there is no net flow of money to or from bankers, to a situation where all that remains is a $1000 tax on bankers. That feels like a tax increase.
Giles views things differently:
One difficulty is that Andrew Bailey, BoE governor, still needs to be persuaded. In 2021 he said the policy would be a tax on banking. The truth is that it would lower public spending.
The “truth” is that truth is a slippery concept, especially where terms are poorly defined. I understand Giles’s point, but I find Bailey’s characterization to be closer to my way of viewing things. You would be essentially forcing banks to lend lots of money to the British government at a rate of zero percent. That seems like the imposition of a tax on banking.
READER COMMENTS
Andrew_FL
Aug 16 2024 at 2:14pm
Only if there are zero administrative costs to the combined policies, which is not realistic.
Thomas L Hutcheson
Aug 16 2024 at 5:30pm
I thought the Fed had gotten rid of IOR? It should it undermines QE when it needs to increase inflation. Whether it is a tax with a DWL or the removals a subsidy that was preventing a DWL, the DWL is the same. Tomato Tomahto.
Craig
Aug 16 2024 at 9:17pm
Hmmm, reserve requirement as a tax? Interesting take, honestly never hears that analogy before. Yeah, makes some sense.
Jose Pablo
Aug 19 2024 at 5:55pm
Well the “tax” would be the interest forgone on the required reserves, but I still don’t see it.
It is a tax only as far as there is a difference between the reserves “required” and the reserves that banks “voluntarily” would hold as a sensible way of making their business.
The voluntary reserves could be roughly equal to (or even higher than) the “required reserves” (at least for some ways of doing this business … maybe the only ways that would survive long term) and in this case, no tax would have been imposed on banks by the reserve requirement.
Neil S
Aug 16 2024 at 11:00pm
To me, this anything but a nothing burger. You are completely ignoring all of the administrative costs associated with these both of these policies. This is a clear net negative for the economy with deadweight costs for both the bankers and the government.
BC
Aug 17 2024 at 2:35am
Funny, my reaction to the first paragraph was that it’s not a nothingburger. It’s an increase in taxing and spending. Among many other objections, my thought was that people would eventually object to the subsidy because bankers are unpopular. On the other hand, any attempt to repeal the tax would be characterized as a tax cut where 100% of the benefits went to bankers. (That’s not speculation on my part. It’s just a recounting of history. Every tax cut is characterized as a “benefit” that goes only to people that previously paid the tax as if somehow never having paid the tax to begin with makes someone worse off because they “miss out” on the tax cut.) I lol’d when, indeed, three paragraphs later this scenario played out exactly as expected.
The other way this could play out would be that other groups would argue that they should also be entitled to the subsidy because bankers shouldn’t get special treatment. The tax would be ignored. An example that’s the reverse of this is how many people have proposed lifting the income “cap” on Social Security (SS). “Why shouldn’t the wealthy have to pay SS taxes on all of their income just like poor people do?” is the argument. These same people never propose increasing SS benefits commensurately with the proposed tax hike.
Another objection is that the policy induces people to try to characterize themselves as “bankers” for purposes of the subsidy but “not bankers” for purposes of the tax. So a non-bank financial institution might argue that it should be allowed to earn interest on deposits at the Fed but not be subject to reserve requirements because it’s not a bank.
As a package, the tax and subsidy might be offsetting. But, once part of the budget, they will be separated whenever people start discussing ways to decrease budget deficits. For transparency, to the extent possible, all taxes and subsidies should be netted out so that, ideally, everyone would either be a pure tax payer that receives no subsidies (“maker”) or a pure subsidy receiver that pays no tax (“taker”). For a given budget deficit or surplus, the best tax and spending policy is the one that minimizes *gross* taxes and spending. That actually matches with Scott’s preference of no IOR and no reserve requirements.
Scott Sumner
Aug 17 2024 at 12:24pm
Everyone, It would be simple to administer the example at the top of the post. Have the Fed announce that all banks would receive a subsidy of $1000, but the subsidy would be taxed at 100%. Then tell the banks that you were doing an income-tax style “withholding system”, and the tax was withheld from the subsidy, so that the banks would receive nothing.
I agree that IOR is more costly to administer, but I never claimed that was a nothingburger. Indeed I oppose IOR, precisely because I think it does have bad side effects.
Ilverin
Aug 17 2024 at 6:36pm
IOR on excess reserves (exceeding the statutory minimum level of reserves) is a subsidy. The narrow bank (defunct), which stores all its deposits at the Fed to receive purely IOR as its entire revenue stream, is the clearest example on how IOR on excess reserves is a subsidy
Scott Sumner
Aug 18 2024 at 1:15am
Maybe, but the post discusses required reserves.
George Selgin
Aug 23 2024 at 12:25am
“This was done because policymakers wished to move to a “floor system’.”
That’s not quite the case. When IOR was instituted no one at the Fed was thinking of a switch to a floor system. The Fed stumbled into that system soon after, as they would have even w/out IOR—for a binding ZLB is effectively a floor system. Eventually they chose to keep to the floor system, which is when IOR really mattered. But they had not planned on that originally.
Comments are closed.