Should central banks adopt a nominal GDP target? | MacroBusiness

Leith van Onselen questions whether the RBA should target a flat growth rate of say 5% for nominal GDP rather than inflation:

I am not convinced that the RBA and RBNZ should necessarily set interest rates around nominal GDP. As shown in the below charts, setting interest rates in this manner would likely see the cash rate rise significantly from current levels which, given anaemic wages growth and high underemployment in both nations, would seem unwise:

Let’s look at the graph of GDP growth a bit closer. If we target 5% GDP growth:

  • From 2001 to 2007 rates were too low. That would have softened the sharp fall in 2008
  • Rates in 2008 were too high
  • Rates were not too low in 2009 to 2010 because of the growth undershoot in 2008
  • Rates were too high 2011 to 2016
  • Again, rates are not too low in 2017 because GDP has undershot its growth target for the last 6 years

I believe that targeting nominal GDP would help to stabilize growth with higher rates in the boom to prevent the need for lower rates in an ensuing bust.

Where I do agree with Leith is that banks need to re-focus from financing largely speculative (housing) assets to financing productive investment. In fact, not just the banks but the entire economy.

Source: Should central banks adopt a nominal GDP target? – MacroBusiness

Fed’s numerical thresholds are a bad idea

The Fed effectively tied its monetary policy to a balloon bobbing in the wind. Pedro da Costa writes on Reuters:

The Federal Reserve on Wednesday took the unprecedented step of tying its low rate policy directly to unemployment, saying it will keep rates near rock bottom until the jobless rate falls to 6.5 percent. That’s as long as inflation, the other key parameter of policy, does not exceed 2.5 percent.

Both unemployment and inflation are moving targets. Unemployment primarily because results are highly dependent on the participation rate: disheartened job seekers who give up looking for work are excluded from unemployment figures. Likewise, inflation measures are highly subjective. Weightings require constant adjustment because of advances in technology and changes in consumption patterns, while cost of housing estimates, which make up 39 percent of core CPI, seem to have little connection with reality. Scott Sumner points out:

The problem seems to be that, according to the Bureau of Labor Statistics, housing prices did not fall. On the contrary, their data shows housing prices actually rising between mid-2008 and mid-2009, despite one of the greatest housing market crashes in history. And prices did not rise only in nominal terms; they rose in relative terms as well, that is, faster than the overall core CPI. If we take the longer view, the Bureau of Labor Statistics finds that house prices have risen about 8 percent over the past six years, whereas the famous Case-Shiller house price index shows them falling by nearly 35 percent…..

Carney broaches dumping inflation target | FT.com

Claire Jones reports that Mark Carney says central banks should consider scrapping inflation targets and target nominal GDP instead — allowing more aggressive measures during a down-turn.

[Mark Carney, next governor of the Bank of England] suggested that a nominal GDP target, where a central bank sets monetary policy based on both inflation and growth, would do more to boost economic output. “For example, adopting a nominal GDP-level target could in many respects be more powerful than employing thresholds under flexible inflation targeting,” he said.

Read more at Carney broaches dumping inflation target – FT.com.

Why the Fed should not target inflation

Scott Sumner, Professor of Economics at Bentley University, proposes that the Fed target nominal growth in GDP (“NGDP”) rather than inflation as Ben Bernanke has long advocated:

“Even he [Bernanke] must be surprised and disappointed with how poorly [inflation targeting] worked during the recent crisis.”

The primary problem, Sumner points out, is that measures of inflation are highly subjective and often inaccurate.

“The problem seems to be that, according to the Bureau of Labor Statistics, housing prices did not fall. On the contrary, their data shows housing prices actually rising between mid-2008 and mid-2009, despite one of the greatest housing market crashes in history. And prices did not rise only in nominal terms; they rose in relative terms as well, that is, faster than the overall core CPI. If we take the longer view, the Bureau of Labor Statistics finds that house prices have risen about 8 percent over the past six years, whereas the famous Case-Shiller house price index shows them falling by nearly 35 percent. That is a serious discrepancy, especially given that housing is 39 percent of core CPI……..

There are errors in the measurement of both inflation and NGDP growth. But to an important extent, the NGDP is a more objectively measured concept. The revenue earned by a computer company (which is a part of NGDP) is a fairly objective concept, whereas the price increase over time in personal computers (which is a part of the CPI) is a highly subjective concept that involves judgments about quality differences in highly dissimilar products.”

Inflation targeting also encourages policymakers to think in terms of monetary policy affecting inflation and fiscal policy affecting real growth — “a perception that is both inaccurate and potentially counterproductive”.

“Advocates like Bernanke see [inflation targeting] as a tool for stabilizing aggregate demand and, hence, reducing the severity of the business cycle. This is understandable, as demand shocks tend to cause fluctuations in both inflation and output. So a policy that avoids them should also stabilize output. I have already discussed one problem with this view: The economy might get hit by supply shocks, as when oil prices soared during the 2008 recession……..”

Linking monetary policy (and the money supply) to nominal GDP growth would offer a far more stable growth path than the present system of inflation targeting.

via THE CASE FOR NOMINAL GDP TARGETING | Scott Sumner (pdf)