Tuesday, June 23, 2015

The Penske View of Macroeconomic Policy


You can learn a lot about macroeconomic policy by driving a Penske truck. I did three years ago when I moved from Texas to Tennessee. The trip began with me driving a 26-foot Penske truck and my wife following in our car. I quickly discovered that Penske had placed a governor on the engine that limited the truck's speed to around 65 miles per hour. This was well below its true potential and made for an incredibly frustrating trip. Hills proved to be especially challenging since I could never generate enough momentum to avoid slowing down as I began ascending them. After the hills it was impossible to make up for lost time because I was prevented from accelerating the truck into catchup speed. There were also the other drivers on the road who got annoyed with my relatively slow speed. One on of those annoyed drivers turned out to be my wife. She decided she could handle the truck better than me so we switched vehicles. She may have done a better job handling the hills and traffic, but only on the margin. Overall, she too faced the same constraint I did: a upper bound on the truck's speed. Reaching our final destination took far longer than we had planned.

There are a lot of similarities between this experience and the use of macroeconomic policy over the past seven years. The U.S. economy, like the truck, has been operating well below its potential. This has meant reaching the final destination of  full employment has taken far longer than anyone expected.

As with the truck, the output gap emerged once the U.S. economy hit a hill called the Great Recession. Unlike the truck, though, this hill was so tough it not only slowed the economy down but caused it to stall and start rolling back down the hill. Fortunately, the brakes were applied, the economy got started again, and the ascent up the hill was made. Since then, the big problem has been the failure of macroeconomic policy to create enough catch-up speed to make up for lost time. More precisely, macroeconomic policy failed to generate enough growth in total dollar spending to close the output gap. 

Macroeconomic policy should have a significant, if not perfect, influence on the growth of total dollar spending through the use of monetary and fiscal policy. The absence of temporarily faster-than-normal catch-up growth in aggregate demand indicates that like the Penske truck there has been a governor placed on macroeconomic policy.

So what is this governor? It is the Fed's 2 percent inflation target. It prevents monetary policy and fiscal policy from generating temporary periods of rapid catch-up growth in aggregate demand because doing so will raise inflation above its target.

For example, imagine that in the third quarter of 2009 the Fed had been able to raise and keep the annualized growth rate of total dollar spending at 7.5 percent until it reached its previous trend (which is near the CBO's full employment level for NGDP). That would be fairly rapid catch-up growth since total nominal expenditures grew on average around 5 percent during the Great Moderation period. Based on historical relationships, this temporary surge in aggregate demand growth would also translate into a temporary surge in inflation.1 As seen below, this inflation surge would last just over two years.


The temporary inflation surge can never occur with a 2 percent inflation target. Yet, as demonstrated by Israel, it was probably needed after the crisis. In terms of the truck analogy, the deceleration of speed on the hill needed to be offset by temporarily faster speed after the hill to maintain  a stable growth path for total dollar spending.

The Penske truck experience also sheds light on two questions that still vex many observers. The first question is why did the Fed's QE programs fail to spur a robust recovery? George Selgin, for example, recently raised this question. The answer is not that these programs are inherently unable to do so. Rather, they were subject to the inflation target governor. Just like I was limited to 65  miles per hour in the truck, the 2 percent inflation target put a limit on how much aggregate demand growth can be generated by the QE programs. Put differently, the 2 percent inflation target meant that the monetary injections created by the QE programs were expected to be temporary. What was needed to spur rapid total dollar spending growth, though, was an expectation that some portion of those monetary injections would be permanent.

To be clear, the inflation target was an upper bound on how much total dollar spending growth could be created by the QE programs. It did not prevent the Fed from using QE programs to prevent a slow down. In fact, it appears 2 percent is actually an upper bound on an inflation target range of 1-2 percent. If so, the QE programs put a floor under the economy even though they were muzzled from spurring rapid aggregate demand growth.

The second question is whether more aggressive fiscal policy could have made a meaningful difference since the crisis. This question is akin to asking whether my wife made a difference when she tried driving the truck. She may have made a difference on the margin, but was ultimately still bound by the governor. Fiscal policy, like monetary policy, is also bound by a governor. It can only create aggregate demand growth up to the point it pushes inflation to the Fed's 2% target. The Fed's preferred inflation measure, the PCE deflator, averaged 1.4% since 2009. That means fiscal policy would have had 60 basis points on average with which to work over the past seven years in closing the output gap. That is not much and nowhere near the two-year run of over 2 percent inflation required to create the catch-up growth in aggregate demand seen in the figure above.2

What policymakers needed after 2009 was a new governor that would have allowed temporary catch-up growth in aggregate demand. The easiest way to have done this is to have  introduced a NGDP level target. Doing so would be similar to replacing the governor on the Penske truck with cruise control. The growth path of total dollar spending would be set and any deviations around that path would be corrected as needed with slowdown or catch up growth in aggregate demand. No matter what your view of the optimal monetary-fiscal policy mix may be, it will be challenging to implement without a NGDP level target. Moving forward,then, this reform is sorely needed. 

I never again want to drive cross country in a truck that has the governor turned on. Similarly, we should never again want to encounter a sharp recession without a NGDP level target.3

1The historical relationship is estimated by regressing the current and two lagged values of nominal GDP growth on the GDP deflator growth.  
2  If you, like Paul Krugman, believe that the Fed actually targets a1-2 percent inflation range then fiscal policy is even more limited since the 60 basis points are gone. 
3There are other reasons to favor a NGDP level target including minimizing the change of hitting a sharp recession in the first place. 

12 comments:

  1. Great post David. I love the Penske truck metaphor. Yes, you need to anticipate a steep climb, and if you fall behind, you want to be able to catch up. I wonder if there's a way to extend the metaphor further, to capture the flaw that inflation targeting is too loose during a positive supply shock. Maybe something about how you don't really want to be going 65 mph when approaching a steep descent, or when approaching a fog bank? Or maybe the metaphor has already done as much for us as it can. Anyway, great post.

    -Ken

    ReplyDelete
  2. David, Maybe now, DeLong will get it! Anyway, some years ago I dubbed the 2% T "Bernanke´s Cross of Gold":
    https://thefaintofheart.wordpress.com/2012/08/19/2-target-bernanke%C2%B4s-cross-of-gold-2/

    ReplyDelete
    Replies
    1. Nice. This point needs repeating until it is better appreciated.

      Delete
  3. David, I just now noticed your artwork in the right margin. Very nice. Have you contacted a gallery? Who's your agent?

    ReplyDelete
    Replies
    1. Tom, very funny. I did those pictures a while back for some blog post, but thought I would bring up front for everyone to see. I have other ideas for economic cartoons, but they take time for me to draw since I am not a natural artist.So my production is limited.

      Delete
    2. David, I really do like those. I prefer the bottom (colored) one to the top, but I think that's only because the top one lacks color. I could see those being published outside this blog.

      Delete
  4. An alternative would be to actually have an inflation target, not an ynflayion ceiling target. In other words a price level trend target . Or as Krugman said central bankers have to promise to be "irresponsible "

    ReplyDelete
    Replies
    1. Thomas,
      Krugman has called for a 4% inflation target for this reason. A higher inflation target or a price level target would solve the aggregate demand shortfall problem outlined in the post. However, for other reasons, I prefer a NGDP level target, primarily that it handles supply shocks better.

      Delete
  5. David, great metaphor. Yet most things in economics are relative. If all the cars had governors at 65 mph, we wouldn't really notice the problem. But we would still remember the days when inflation was harder to control.and the governor was needed to keep inflation below the speed limit.
    A better metaphor would use a plane that has to rev up faster to avoid stall speed and then to return to cruising altitude. A plane would keep falling to lower altitudes using your governor.

    ReplyDelete
  6. Jose Romeu RobazziJune 25, 2015 at 7:37 AM

    Prof. Beckworth
    Do you have a suggested target for NGDPLT, should it be adopted ? Do you believe that the implicit inflation level embedded on it matters for labor market efficiency ? Thank you.

    ReplyDelete
    Replies
    1. Jose,
      I think the biggest implications of NGDPLT for labor are (1) it would tend to stabilize nominal income growth, (2) it would allow the price level to move around more, (3) as a consequence of (1) and (2) it is that it would better align real wages with productivity growth. See some related thoughts here: http://macromarketmusings.blogspot.com/2015/04/a-partial-solution-to-income-inequality.html

      Delete
  7. Ken and Edward,
    Yes, we could extend the analogy along the lines both of you suggest. My hope is simply to get people thinking about the "macro space" in which monetary and fiscal policy operate and to see its limits under a 2% inflation target.

    ReplyDelete