$5.5 trillion, the rest is irrelevant. But we can ask all the questions of Section 14.1 about helicopter drops: Suppose the M2 expansion had been entirely produced by purchasing existing Treasury securities, with no deficits. Would this really have had the same effect? Are people starting to spend their M2 because they don't like the composition of their portfolios, which have too much M2, paying 0.01% (Chase), and not enough one-year treasurys paying 0.1%? Or are they simply spending extra ``wealth? Suppose the Federal Reserve had refused to go along, and the Treasury had sent people Treasury bills directly. Would that not have stoked the same inflation? The fiscal expansion, the wealth effect, is the natural interpretation of this episode.
Why did this stimulus cause inflation, and that of 2008, or the deficits from 2008 to 2020, did not do so? Federal debt held by the public doubled from 2008 to 2012, as inflation went nowhere. (Figure 5.6.) From the fiscal theory point of view, the key feature is that people do not believe this debt will be paid back. (In principle discount rates or real interest rates could be different, but in this case they are nearly the same.) So why, this time, did people see the increase in debt and not infer higher future surpluses, while previously they did? Several speculations suggest themselves.
First, we can just look at what politicians said. In 2008 and following years, the Administration continually offered stimulus today, but promised deficit reduction to follow. One may take those promises with a grain of salt. But at least they bothered to try. This time there was no talk at all about deficit reduction to follow, no policies, no plans, no promises about how to repay this additional debt. Indeed, long-run fiscal policy discussion became focused on how low interest costs, r<g, and modern monetary theory allow painless fiscal expansion. The discussion of tax hikes in the spring and summer of 2021 focused entirely on paying for some of the ambitious additional spending plans, not repaying the Covid helicopters.
Second, much of the expansion was immediately and directly monetized and sent to people as checks. The previous stimulus was borrowed, and funded government spending programs that have to gently work their way into people's incomes. Following 2008, M2 did not rise as much as debt. The QE operations were mostly confined to a switch of bank assets from Treasury debt to reserves, as we see from the contrast between the monetary base (currency + reserves) and M2 in Figure 2. (Note the base is on a different scale.)
Money and bonds may be perfect substitutes, but who gets the money or debt and how can still matter. Traditional buyers of Treasury debt have savings and investment motives. (Think of an insurance company.) If the Fed instantly buys the debt, and Treasury sends reserves (checks) to people, the larger debt goes quickly to people who are likely to spend gifts quickly. Debt sold to traditional bond purchasers, who show up at Treasury auctions or buy from broker-dealers, sends a different signal than money simply created and sent to people. This statement implies some slow-moving capital frictions and heterogeneity, but most of all it echoes the idea that the institutional context of debt expansion matters to expectations of its repayment. We distinguished Treasury actions as a share issue and reserve creation as a split, differing only in the expectations of repayment that each engenders. We distinguished ``regular budget and ``emergency budget deficits, likewise signaling backed vs. unbacked expansions. Since the desire was stimulus, and stimulus requires the government to find a way to communicate that the debt will not be repaid, one can regard the effort as a success at its goal, finally overcoming the expectations of repayment that made previous stimulus efforts fail, guilty only of having overdone it.
Third, the Covid-19 economic environment was clearly different. The pandemic and lockdown shock is fundamentally different from a banking crisis and recession shock of 2008. GDP and employment fell faster and further than ever before, and then rebounded most of the way, also faster than ever before. From a macroeconomic point of view, the Covid-19 recession resembles an extended snow day rather than a traditional recession. One might call that a ``supply shock, as the productive capacity of the economy is temporarily reduced, but demand falls as well, for the same reasons that people don't rush out to buy in a snow day either. (I write ``from a macroeconomic point of view. A million people died in the U.S., an ongoing public-health disaster. Many people suffered lost jobs and businesses.) Roughly speaking the economy was operating at its reduced ``supply capacity all along, not needing extra ``demand. Providing that ``demand could reasonably spill more quickly to inflation.}}