In this note, we use FRB-US to show that the US debt/GDP ratio becomes explosive going forward under several scenarios, and we discuss the implications. First, we explain the (surprising) implicit assumptions in the latest SEP-consistent database. Then, we run a set of scenarios, including fiscal adjustments, using the “inconsistent/unconstrained FRB-US model” to assess how the relevant variables (deficit/GDP, debt/GDP, FF rate, and yields) can evolve going forward. We show that the SEP-consistent database is based on unrealistic assumptions: the deficit/GDP ratio returns quickly to low levels because the average tax rate (on personal income) increases notably and long-term rates are assumed to be way lower than the current levels. On the other hand, when the model is run unconstrained (and updated to reflect current yields), it is very difficult to stabilize both the deficit/GDP and the debt/GDP ratios even assuming credible fiscal adjustment plans. In most scenarios, the FF rate is projected at (at least) 4.5% at the end of 2024, unless Congress passes a large increase in taxes and a permanent cut of government spending, which are both unlikely. We provide the path of the 5y, 10y, and 30y yield implied by each scenario, including assuming a recession in 2024. To sum up: despite the rhetoric of the (supposed) success of “big fiscal”, the truth is a cold one. The fiscal situation of the US is worrying, as we might be at a “tipping point” where the debt dynamics becomes explosive. Right now, it is unclear how the US can achieve fiscal sustainability in the medium-term, given the size of the issue and the lack of political commitment.
If the reader has questions, please direct them to our FRB-US specialist: Tilda Horvath (tilda.horvath@underlyinginflation.com). Tilda has programmed FRB-US at the Board and managed the model for almost 20 years. We remind the reader that we run ad-hoc scenarios using FRB-US free of charge. Do not be shy.
How fiscal policy “works” in FRB-US
Note: the latest run of FRB-US, including the current baseline, is here. In this note, we follow the color convention of the previous FRB-US notes: blue lines refer to the SEP-consistent, red lines to the current baseline (the “inconsistent” FRB-US model), and green lines to alternative scenarios.
The SEP-consistent database is run under a non-credible “deficit targeting” rule. In the “unconstrained” FRB-US model, the deficit does not stabilize. The SEP-consistent database published by the Fed Board replicates the forecast of the variables shown in latest Summary of Economic Projections (SEP). While fiscal policy is not the main spot of interest of the model, FRB-US contains a fiscal block. The key point (generally not reported) is that the SEP-consistent dataset is run under a “deficit targeting” rule, that is imposing to reach a desired deficit/GDP ratio “at some point in the future”. Specifically, the latest database is constructed imposing a deficit/GDP target of 2.7%. Consequently, in the latest SEP-consistent dataset (blue line in Figure 1), the deficit/GDP ratio returns pretty quickly to moderate levels by construction and reaches the desired target at the end of 2028. This scenario is way more optimistic than the latest CBO projections (here) because in the SEP-consistent database the average tax rate on personal income increases by almost 2% in each year of the forecast horizon and yields remain very low (for instance, the 10y yield is projected around 3.6% by the end of 2023 – see blue lines here). On the other hand, when the model is run “unconstrained” (red line in Figure 1), the deficit/GDP ratio does not stabilize because the interest payments become too large (reminder: in the current baseline (red line), the 10y yield averages 4.7% in 2023:Q4 and remains around 4.8% until the end of 2024. We remind the reader we can run alternative scenarios based on his/her own assumptions). Consequently, while in the SEP-consistent database the debt/GDP ratio (Figure 2) stabilizes in 2026, in the “inconsistent FRB-US”, that is the more realistic version, the dynamic becomes explosive.
(Note: in FRB-US the “debt” is defined as “federal debt held by the public”)
Figure 1. Federal government fiscal balance to GDP ratio. (blue-SEP; red-model baseline)
Figure 2. Federal debt (held by the public) to GDP ratio. (blue-SEP; red-model baseline)
Scenario #1. The “inconsistent deficit targeting”
Targeting a deficit/GDP ratio in the “inconsistent FRB-US” does not solve the problem and the debt/GDP ratio remains explosive. So far, we have shown that the dynamics in the SEP-consistent database are non-explosive because the deficit is forced to drop quickly and interest rates are way lower than current levels. The first natural question is then: what happens if we impose a deficit target to the “inconsistent FRB-US”? (Reminder: the SEP-consistent database is constructed by the Fed Board staff by fitting “add factors” to relevant variables so that they fit the latest SEP. The “add factors” are, literally, factors that are added to each variable so that the forecast matches the SEP). Figure 3 shows the deficit-targeting scenario (the green line) compared to the “inconsistent” baseline (the red line). In this scenario, we asked the model to reach a certain deficit target without specifying any preference in terms of policy mix (lower G vs higher T); in part II of this note, we show the model forecast under more specific fiscal shocks. The result of this scenario is that when the model is forced to reach a deficit target (of 2.7%), it does so very gradually (the green line in Figure 3) and by raising the average tax rate on personal income as in the SEP-consistent database. However, even in this scenario, because the deficit/GDP is large and yields are high, the debt/GDP does not stabilize and becomes explosive (see here).
In other words, as we wrote above, in the SEP-consistent database, the debt/GDP ratio stabilizes because of the combination of two factors: the deficit reaches reasonable levels quickly and long-term interest rates are assumed way lower than current levels. A corollary of this first scenario exercise is that at this point, it is enough to take out one of those two factors for the debt/GDP ratio to become explosive.
Figure 3. Federal government fiscal balance to GDP ratio. (red-model baseline; green-scenario)
Targeting a deficit/GDP ratio produces lower growth, higher unemployment, lower inflation, and a lower path of the FF rate. Figure 4 shows the evolution of the main variables of interest under the “inconsistent” model baseline (red lines) and the deficit-targeting scenario (green lines). Because taxes increase in the deficit-targeting scenario, real GDP growth is projected to be even more anemic than in the current baseline. Consequently, the evolution of the unemployment rate is less favorable and it goes well above 5% by the end of the forecast horizon. Core inflation is less persistent but remains above target. The monetary authority starts cutting rates in 2024:Q2: the FF rate is projected at 4.8% at the end of 2024 before falling to 2% by the end of the forecast. The evolution of the 5y, 10y, and 30y yield implied by this scenario can be seen here. Under the deficit-targeting scenario, the 10y yield remains elevated, although mechanically lower than in the baseline.
Figure 4. Evolution of main variables. (red-model baseline; green-scenario)
Note: Real GDP growth and core inflation are expressed as YoY. Core inflation is core PCE price inflation. The red line shows the “inconsistent FRB-US” forecast (the current baseline), that is the model-based forecast removing the “add-factors” put by the Fed staff to match the latest SEP. The green line shows the deficit targeting scenario.
Conclusion of part I and what to expect in part II
In part I, we have shown that the SEP-consistent database contains very optimistic assumptions about the evolution of the fiscal deficit. In a more realistic scenario, the deficit/GDP ratio remains elevated and the debt/GDP ratio is explosive. (In a sense, this should not come as a surprise, given that the latest CBO projections contain a similar result, even if the CBO assumptions about long-term yields are quite generous, given what happened recently). In part II, we show how FRB-US expects the US economy to behave under different scenarios: (i) a one-time exogenous 1% shock to federal purchases, (ii) a combination of a gradual increase in the average tax rate and a sustained spending cut on government purchases, and (iii) a negative demand/recessionary shock. The main conclusion of part I holds under those scenarios. The overall conclusion is pretty worrying: when looking at the next 10 years or so, it is very difficult to see a world in which the debt/GDP ratio stabilizes. Given the size of the issue, the stabilization of the debt/GDP ratio would require a very committed fiscal authority, a circumstance which appear unrealistic at the moment.