The NY Fed has published two blog posts (Drivers of Inflation: The New York Fed DSGE Model’s Perspective and Disinflation Policies with a Flat Phillips Curve) which analyze two separate questions: (i) the drivers of current elevated inflation, and (ii) how alternative monetary policy strategies might bring inflation back to target. Both questions are examined through the eyes of the NY Fed DSGE model. DSGE models are especially useful tools to address what factors underlie the recent surge in inflation because their structure enables to decompose the evolution in terms of underlying driving forces. Not only, but DSGE models can be used to design and assess the optimal monetary policy response in face of different types of shocks.
To keep in mind
According to the NY Fed DSGE model, the current inflationary wave is driven by persistent cost-push shocks. As such, inflation is expected to remain elevated this year and slowly converging close to (although still above) target in 2024. Importantly, monetary policy faces an unfavorable trade-off when attempting to stabilize inflation in response to cost-push shocks, due to an extremely flat Phillips curve. Lowering inflation requires a deep and protracted contraction, regardless of the policy strategy underlying the pursuit of this objective.
What the papers do
Using the NY Fed DSGE model (documentation here) the authors decompose the surge in inflation in five groups of shocks: 1) cost-push shocks that hit the economy before 2020; 2) cost-push shocks that hit the economy in 2020; 3) cost-push shocks that hit the economy in 2021; 4) transitory price level shocks; and 5) COVID-19-related shocks (defined as an i.i.d. shock, see paragraph 2.3 at page 8 of the model documentation).
Cost-push shocks are defined as shocks to the price Phillips curve that determines inflation in the model (i.e. shocks to the price of energy or other intermediate inputs, including those due to bottlenecks in the supply chain), as opposed to shocks that move inflation through their effect on marginal costs (productivity, real wages, and the cost of capital). The key distinction between cost-push shocks and transitory price level shocks in the NY Fed DSGE model is that the effects of the former on inflation are persistent. By including both transitory and persistent pricing shocks, the model encompasses two key narratives that have animated the debate on the sources of inflation over the past year (“transitory” vs “persistent”).
In the second paper the authors study how alternative monetary policy strategies might contribute to bringing inflation back down to 2 percent. The authors consider two monetary policy strategies (FIT = flexible inflation targeting, and FAIT= flexible average inflation targeting), and two types of shocks (cost push shocks vs aggregate demand shocks).
Results
Starting with the contributions to high inflation, the NY Fed DSGE model points in the direction of persistent inflation. The chart below shows the main result (figure is in deviation from 2 percent) and the model forecast. The main takeaway is that according to the model cost-push shocks have driven inflation steadily higher over the course of 2021 and are expected to fade only gradually over the subsequent two years.
Transitory shocks dominated the movements of inflation early in the pandemic, as it first plunged in the second quarter of 2020 and subsequently rebounded in the third quarter as the economy shut down and then re-opened. Persistent cost-push shocks were virtually absent in that first phase of the pandemic economy. The picture changes drastically in 2021, when the model attributes about 2.5 percentage points of the surge in inflation to new cost-push shocks. The temporary price level shock contributes another 1.5 percentage point to inflation in the second quarter, with a much smaller contribution in the third quarter. But this moderating effect is swamped by the size and persistence of the cost-push shocks hitting in the first half of the year. These shocks dominate the current inflation picture, imparting significant persistence to the forecast.
How can a monetary authority react? Switching to the monetary policy response, in the second blog post the authors main finding is that there is no monetary silver bullet. The intuition is that because the Phillips curve is still estimated flat in the model, monetary policy can only achieve faster disinflation at a considerable cost in terms of forgone economic activity. This is true regardless of the systematic approach followed by the central bank in the model to pursue its objective. The figure below shows the IRFs of the main variables of interest to a cost-push shock that leads to a surprise increase in inflation of roughly 1 percent (QoQ a.r.).
The response of inflation to the cost-push shock is very similar regardless of the policy rule in place. The FAIT rule reduces inflation a bit faster than the historical FIT rule, but this small benefit in terms of inflation control comes at a very large real cost. The recession that monetary policy must engineer to achieve this negligible reduction in inflation is deeper and more protracted under the FAIT than the FIT policy, as judged by any of the real variables in the model: hours worked, GDP growth, the level of GDP, and (real) marginal costs. This very unfavorable sacrifice ratio reflects the extreme flatness of the estimated Phillips curve. FAIT “gets the job done” with minimal movements in the policy rate because the stance of policy is expected to remain tight for (much) longer than under the estimated FIT rule. In other words, the nominal FFR (top right panel above) react less on impact with FAIT than with FIT but remain higher for much longer.
This tighter-for-longer policy implies that the 10-year average expected interest rate rises less on impact than under FIT, but declines more slowly. It also implies that the contraction in economic activity is more protracted, to the point that long-term inflation expectations actually fall, in spite of the inflationary shock. The increase in long-term nominal rates and the fall in inflation expectations result in a larger increase in real long-term rates under FAIT than under FIT, making the former rule more contractionary than the latter, even if it implies a smaller increase in the nominal short-term interest rate. This explains why the recession is more severe, and lasts longer, under FAIT than under FIT.
Comment
Very interesting analysis that confirms the narrative of the incoming data.
According to the NY Fed DSGE model, the current inflationary wave is driven by persistent cost-push shock. As such, inflation is expected to remain elevated this year and slowly converge close to (although still above) target only in 2024. Monetary policy faces an unfavorable trade-off when attempting to stabilize inflation in response to cost-push shocks, due to an extremely flat Phillips curve. Lowering inflation requires a deep and protracted contraction, regardless of the policy strategy underlying the pursuit of this objective, although the contraction is more pronounced under FAIT.
Implications for the Fed
So far, the Fed has been careful with the conduct of monetary policy to minimize the risk of crashing the economy. However, the NY Fed analysis shows that under FAIT the cost to disinflate the economy is larger than under FIT (while delivering similar inflationary paths) if inflation is generated by persistent cost-push shocks. Therefore, should inflation remain elevated in the next few months, the Fed will have a necessity (and a rationale) to be more aggressive than it has been.
The data will dictate the speed and aggressiveness of monetary policy, as usual.
But ultimately, we might.. go back to the 50s(bps) sooner than expected.