"Make America Keynesian Again" part 2
In yesterday's blog post I wrote about why I believe it is the combination of Donald Trump's fiscal stimulus plans (infrastructure investments and tax cuts) combined with the Federal Reserve's willingness not to (fully) offset this, which has pushed inflation expectations in the bond markets up very significantly since Tuesday. If the Fed's inflation target was fully credible fiscal stimulus would be fully offset by the expectations of a tightening of monetary policy to "neutralize" the impact on aggregate demand from fiscal stimulus. This of course is known as the so-called Sumner Critique. I would normally think that the Sumner Critique would hold and announced fiscal stimulus or fiscal contraction would not impact inflation expectations. This is for example what I argued in 2012 and 2013 in relationship to the so-called fiscal cliff (see here, here and here). That argument of course turned out to be completely right - the fiscal contraction did not cause inflation expectations to drop and the US economy did not fall into recession contrary to what was argued buy arch-Keynesians such as Paul Krugman. However, as I have often argued the causality in the economy as well as the impact of fiscal shocks depend critically on what kind of monetary policy rule the central bank has (see fore example here, here, here and here). The standard textbook example is the Flemming-Mundell model where the budget multiplier is zero in a free floating exchange rate regime, but positive ("keynesian") in a fixed exchange rate regime. The important point in relationship to the expected fiscal easing from the Trump administration is that the Federal Reserve explicitly have called for the kind of fiscal stimulus that Trump now wants to deliver meaning that the Fed effectively have signaled that they will not fully offset the impact on aggregate demand. Furthermore, it is clear that the Fed has a preference for higher nominal interest rates - disregarding the level of inflation expectations. The Fed simply don't like interest rates at this level. However, the Fed also realizes that it is not really possibly both to deliver higher inflation than presently (which is necessary to hit the 2% inflation target) and increase interest rates. But they can get both by allowing fiscal policy to increase aggregate demand. This, however, necessitates that the Fed will not increase interest rates quite as fast as the rise in the equilibrium interest rates caused by easier fiscal policy. This means that the Fed effective will need to peg the interest rate level. Trumponomics in A simple IS/LM model with two different policy rules These consideration have made me think about how to illustrate this in a simple model that even first-year economics students can understand. That model is a rudimentary IS/LM model. While drawing with my 6-year old son tonight I put the equations on a paper (yes, I know am sometimes a nerdy dad...). Here they are: What we have here is two equations. One for aggregate demand (AD) and one for money demand as well as a monetary policy rule - or rather three different monetary policy rule. Equation (1) simply says that aggregate demand is composed of private spending/investment, which dependent on the interest rate level (r) and of government "spending". Higher interest rates causes private spending/investment to drop. Equation (2) is a standard textbook money demand function, where money demand (m) depends on nominal GDP (P*Y) and the interest rate level. Higher interest rates causes money demand to drop. In the standard IS/LM model we use this to construct the LM and the IS curves. However, we also need some monetary policy rules. Introducing a monetary policy rule is what I earlier has termed a IS/LM+ model (See here and here). The two policy rules for the Fed I here look at is an interest rate target rule - (3)' - and a nominal GDP target rule - (3)''. In the interest rates targeting case we simply assume that the Fed will increase (decrease) the money base (m) if the interest is higher (lower) than the interest rate target (rT). If the coefficient lambda is set to be equal to infinity it means that the Fed will not accept any change in the interest rate from the target. Our nominal GDP target rule (3)'' essentially works in the same way. If nominal GDP is below (above) the target then the money base is increased (decreased) to push up (down) nominal GDP. We can also illustrate this with graphs - again a bit from the kitchen table: Lets first start in the standard IS/LM model. Here fiscal easing - higher g in the model and in real-life it is Trump's tax cuts and infrastructure investments - causes the IS curve to shift to the right. This pushes up nominal GDP. In fact in the textbook prices are assumed to be fixed so P=1. We don't have to make that assumption here. The increase in g also push up the interest rate (r) because it in the standard IS/LM model is assumed that the money base (m) is fixed. In the graphs above this is the move from 1 to 2. The NGDP rule - full crowding out of Trump's fiscal easing However, if we have an NGDP rule we will see that nominal GDP (PY) has now been pushed above the NGDP target. This will cause the Fed to reduce the money base (m) and the Fed will continue to reduce the money base until NGDP is back on target. This causes the LM curve to shift to the left - the the lower graph that is the shift from 2 to 3 causing a further increase in interest rates. This increase in interest rates will - see equation (1) - cause private spending/investment to drop exactly as much as government spending/investment (g) has increase. Said in another way we have full crowding out and the budget multiplier is zero. In this case the Sumner Critique obviously applies. Therefore, if the Fed follows a NGDP targeting rule then this model tells us that Trump's infrastructure investments will just crowd out private consumption and investment and hence not create the millions of jobs he has promised and it will not increase inflation. The interest rate rule - Trump's boom (and bust) However, since Tuesday we have seen inflation expectations increase significantly. Just take a look at 5-year/5-year swap forward inflation expectations. Long-term inflation expectations are up more than a quarter of a percentage point since Tuesday morning. In fact this is the largest two-day increase in inflation expectations since April 2015. This is certainly not a small change in inflation expectations. Therefore the markets are telling us that we should not expect full crowding out of Trump's fiscal easing. Lets turn to the explanation - interest rate pegging. This is what we have in the upper graph. Trump eases fiscal policy. This pushes the IS curve from 1 to 2. This increases (nominal) GDP growth and push up interest rates. However, if the Fed effectively has an interest rate rule then it will not try to offset the increase in GDP, but rather will try to offset the increase interest rates by increasing the money base (rather than reducing it). This is the shift in the LM curve to the right in the upper graph, which causes nominal GDP to increase further. Obviously in real-life the Fed will not keep the interest rate completely fixed, but it might choose to increase interest rates less than the increase the equilibrium rate caused by massive fiscal stimulus. This sounds like something out of the 1970s' insane "keynesian" policy mistakes, but I actually think it is pretty much what Janet Yellen would like to see. This is what she said a few weeks ago:
If we assume that hysteresis is in fact present to some degree after deep recessions, the natural next question is to ask whether it might be possible to reverse these adverse supply-side effects by temporarily running a "high-pressure economy," with robust aggregate demand and a tight labor market.Effectively Yellen is saying she would like to see the US economy "overheat". Trump would like the same thing and the markets understand that. The coming conflict between Stanley Fischer and Donald Trump I overall think that these very simple models pretty well discuses the connection between monetary policy (rules) and fiscal policy and how different rules can significantly impact how the US economy response to Trump's planned fiscal stimulus. And I am mostly inclined to think that the Fed will implicitly collude with the Trump administration to create exactly the kind of high-pressure economy that Yellen was talking about and hence in the next couple of months we might want to think about the US economy as the interest rate targeting case in my model. However, it is also very clear that anybody who remembers the 1960s and particularly the 1970s knows that this could be an extremely dangerous strategy. In fact Fed Vice Chairman Stanley Fischer has already warned against it. This is what Fischer said a couple of weeks ago:
“If you go below the full employment rate, or peoples’ estimates of full employment, by a couple of tenths of percentage points, I don’t think there’s any danger in that...But saying we should keep going until the inflation rate shows us we’re wrong, then you’re going to change too late.”Said in another way Fischer might be willing to go along with keeping interest rates below the equilibrium interest rate for some period, but he clearly fear that such a strategy soon could cause inflation to spike. And I would certainly agree with him. Monetary policy in the US has certainly more or less consistently been too tight since 2008, but we have recently moved towards a more neutral monetary policy stance and the combination of Yellen's ideas about a "high-pressure economy" and Trump's fiscal expansion could be what pushes inflation expectations significantly above 2%. If that where to happen I would expect Yellen and the Fed to reverse cause and start tightening monetary conditions rather aggressively. Donald Trump certainly would not like that and that might be setting us up for a conflict at some point in 2017 or 2018 between the Federal Reserve and the Trump administration.