Don't mix fiscal and monetary policy

Today’s post is brought to you by my sponsor, Mechanize. They’re hiring junior software engineers at $300K/year base salary. Apply now!

* * *

[To head off questions about “What should the Fed do this week?”, consider the fact that financial markets currently do not seem to be anticipating excessively low inflation or excessively high unemployment going forward. (Admittedly, it’s hard to be certain.) At the same time, financial markets are anticipating that the Fed will raise its interest rate target. That suggests that a rate increase would not be inappropriately hawkish.]

Over the past decade or two, I’ve seen a dramatic increase in discourse that attempts to link fiscal and monetary policy. This is unfortunate, part of the more general decline in the field of economics since 2008. I see two particularly common mistakes, which I’ll consider one at a time:

1. The false view that monetary policy has important fiscal implications for a country like the US.
2. The false view that stabilization of aggregate demand should be done with a mix of monetary and fiscal policy tools.

Back in the 1990s and early 2000s, the profession had achieved a consensus that monetary policy was the appropriate tool to target aggregate spending, and that fiscal policy should aim at other objectives such as encouraging long run growth, providing public goods and redistributing income. Unfortunately, that consensus is gone.

Part 1: Monetary policy, seigniorage and the real value of the public debt

Monetary policy certainly does have some fiscal implications, in two primary areas. First, the Fed earns a profit from issuing zero-interest currency. You can think of that profit in terms of the difference between the face value of a “Benjamin” (i.e., a $100 bill), and the cost of printing that $100 bill, which is about 11 cents. More often, however, people view the profit in terms of interest earned on the Treasury securities held on the asset side of the Fed’s balance sheet, which were purchased when the currency was issued. (These two approaches are analogous to valuing a share of stock in terms of either its market price or its expected flow of dividends.)

Today, the stock of currency in circulation is roughly $2.5 trillion. If the Treasury securities held by the Fed earn 4% interest, then the Fed earns about $100/year in easy risk-free profits from its currency monopoly. It’s like a $2.5 trillion hedge fund that borrows at 0% and lends risk-free at 4%.

Until 2008, the flow of profits to the Fed was fairly stable, as the monetary base was roughly 98% currency. After the Fed began paying interest on reserves, things got a bit more complicated. On average, it remains true that almost all of the Fed’s seigniorage comes from the zero-interest currency. But as with the man who drowned in a lake that averaged 3 feet in depth, averages can be misleading.

The base is now over 50% composed of commercial bank deposits held at the Fed, and those deposits (i.e., bank reserves) earn interest roughly equal to the rate of interest on short-term Treasury debt. That means that the Fed will earn either a loss or a profit on the reserve portion of the monetary base depending on whether the short-term interest rate (i.e., IOR) is above or below the rate earned on the Treasury’s holdings of longer-term bonds. (The Fed may also earn capital gains or losses when Treasury debt is sold.)

On average, these reserve-based interest inflows and outflows will be roughly a wash, but as you see from the graph below from a PIIE paper by Asher Rose), the Fed’s profits will be unusually high when short-term rates are below the rate earned on their existing stock of T-bonds (as during the 2010s), and unusually low during periods where the short-term rate is above the rate earned on the Fed’s stock of longer-term bonds, as during 2023-24. But the average profit won’t be greatly affected by the Fed’s 2008 decision to start paying IOR.

I mentioned that you might expect the Fed to earn about $100 billion/year in seigniorage from the $2.5 trillion currency stock, on average. That flow of income is roughly 0.3% of GDP and represents the Fed’s normal contribution to funding federal spending (which is currently about 23% of GDP.) That’s not nothing, but it is a minor contribution. Since 2008, the flow of seigniorage has become more volatile, but not enough to have important fiscal implications. It’s small potatoes. The Fed should just focus on stabilizing NGDP and ignore the effect of monetary policy on the government’s fiscal situation.

BTW, I often get accused of advocating erratic monetary policy when I point to thought experiments involving doing “whatever it takes” at the zero bound. Exactly the opposite is the case. A policy of NGDP level targeting would have resulted in a much smoother path of seigniorage that what you see in the graph above. The outsized profits of the 2010s and the outsized losses of the early 2020 were caused by policies that led to highly erratic NGDP growth. Don’t conflate thought experiments aimed at making a theoretical point with the likely path of policy under NGDP targeting. Under NGDP level targeting, there are far fewer instances of the zero lower bound. Interest rates become more stable.

None of the discussion above means that monetary policy can never have important fiscal implications. Hyperinflation can lead to significant seignorage. And the gradual one-time shift from a gold standard to fiat money did reduce the real burden of the federal debt. But under a 2% inflation targeting regime, it is not worth thinking about the fiscal implications of monetary policy—it’s just not that important.

People also focus too much on interest rates, as if the government sets them with a magic wand. In fact, the real interest rate is almost entirely determined by market forces in the long run, not by monetary policy. The Fed can influence nominal rates in the long run, but only by changing the trend rate of inflation. And there are two problems with trying to affect the fiscal situation through inflation.

First, the public hates inflation, much more than they hate taxes. The recent inflation was far more unpopular than the recent rise in tariffs, although neither are particularly popular. The public would not like a 10% VAT, but they’d vastly prefer a 10% VAT to the sort of hyperinflation that would be required to raise an equal amount of revenue through seigniorage.

Second, even if the pubic did accept modestly higher inflation, it would do little to address our fiscal problems. A 1% higher annual rate of inflation reduces the real value of government bonds by an extra 1%/year, but it also increases the interest cost of the public debt by the same 1%, due to the Fisher effect. It’s a wash. At best, there is a one-time gain from an unexpected transition to higher inflation, but we’ve already done that several times. We are already paying a price for the public’s skepticism about the government’s willingness to hold inflation down, due to previous policy mistakes. Sorry, monetary gimmicks aren’t going to address our fiscal problems.

Before addressing the second monetary/fiscal fallacy, I’d like to briefly discuss a previous Fable response to my discussion of the Great Recession:

The fiscal foundations of the monetary anchor. Sumner's newest claim — no fiscal constraint hinders credibility when inflating — meets a literature the series never names. Del Negro–Sims (2015) and Hall–Reis (2015): a central bank that expands with long-duration assets while paying interest on reserves faces remittance losses and possible negative equity when it later tightens; absent a fiscal indemnity, anticipation of that state and its politics constrains the willingness to promise inflation. That is a manufactured "won't" — the series' own category. Exhibits: the SNB's 2013 loss forced it to skip its distribution to the cantons the year before the floor fell; the Fed has carried a deferred asset past $200 billion since 2022; the Bank of England's asset purchases run under an explicit Treasury indemnity — armor done as fiscal engineering. This cuts at the synthesis itself: "anchor plus stabilizers" treats the anchor as purely monetary, but a credible whatever-it-takes anchor requires fiscal underwriting at its foundation. The armored regime is a fiscal-monetary treaty; the fiscal authority is present at the creation even in the market monetarist first-best.

Fable is making far too much of a point that, while theoretically valid, is of little practical importance. Keep in mind that the Fed is part of the federal government’s consolidated balance sheet. That means that any loss to the Fed from a fall in the real value of its Treasury bonds is exactly offset by a gain to the Treasury, as the real value of its future tax obligations falls by an equal amount. Even in the highly unlikely event that the Fed might someday require a “bailout”, it is not a problem worth worrying about.

The demand for Swiss francs is somewhat larger and more volatile than the US dollar (as a share of GDP.) That’s partly due to the Swiss franc’s status as a safe haven currency within Europe, and partly due the the Swiss decision to target inflation at an unusually low rate. But Switzerland is a rich country and can easily afford to self-insure any volatility in central bank income from what in the long run will be a significantly positive flow of income from seigniorage. This is not a “fiscal problem” worth worrying about.

The Asher Rose article I linked to above is excellent, but I’m going to quibble with the final paragraph:

These losses do not impair the Fed’s ability to conduct monetary policy. The central bank, unlike a traditional corporation, can lose money for a sustained period. But in an increasingly politicized age, optics matter. At a time when trust in institutions is already fragile, public misunderstanding of these losses could complicate communication and erode support for the Fed’s independence.

Rose is expressing the conventional wisdom, but is this true? Go to your local shopping mall and ask random people if they are losing sleep over the Fed’s losses during 2023-24, or indeed whether they even knew about them. I rarely meet people that even know what monetary policy is. They ask me: “Monetary stimulus? Is that sort of like when the government gave people checks?” Sigh . . .

I worry that a balanced fiscal/monetary approach is increasingly seen as the wise approach, an indication that you aren’t some sort of nutty extreme monetarist that monomaniacally focuses on money and ignores the fiscal perspective. In the next section, I’ll show why the sensible pragmatists are wrong; monetary and fiscal policy need not be “coordinated”.