2011-10-27

Central bank transparency

Nick Rowe has very good post on central bank transparency.
One sentence in Fed vice-chair Janet Yellen's speech caught my eye:
"Indeed, I believe that the Federal Reserve qualifies as one of the most transparent central banks in the world."
That is total bullshit. The Fed is one of the least transparent central banks in the world.
The complaint Nick has is that the goal of the Fed is very far from transparent and that this is the most important area for a central bank to be transparent in. The reason for this is that market expectations of future monetary policy has a huge impact on the economy today. If the central bank refuses to reveal its goal, the market is forced to guess by carefully reading anything the central bank releases. A good example of how this guessing game can cause bad effects on the economy can be found in a recent Bloomberg article.
At BNP Paribas SA’s New York trading desk, Julia Coronado, the bank’s chief North America economist, watched as three words helped undermine the Federal Reserve’s latest attempt to aid the U.S. economy: “significant downside risks.”
The phrase, tucked into a seven-paragraph policy statement about the Fed’s plans to move $400 billion into long-term debt from short-term bonds, warned about the economic outlook while offering no clue on the risks’ severity. The Sept. 21 statement said Fed officials expected “some pickup” in the pace of recovery, though unemployment would “only gradually” decline.
Bloomberg continues.
Without such specifics, Fed watchers like Coronado, who keeps a rubber Bernanke doll taped atop her computer screen, are left to parse the qualitative language of Fed statements. Last month, investors decided the words meant it was time to buy bonds and sell stocks; the Standard & Poor’s 500 Index closed 3 percent lower that day.
This isn't directly about the Fed's goal. The market wanted to know exactly what the downside risks were and why the Fed had reached that conclusion. What model had they used, and so on. It still shows how big effect three simple words from the Fed can have. How much easier wouldn't it be if the Fed's goal were clear? And if they also published all the details for their predictions, well, that would be the icing on the cake.

The Bloomberg article also has a nice little tidbit from Bernanke.
He has argued in favor of providing “quantitative guidance” on the Fed’s inflation goal and once criticized central banks that engage in “Marcel Marceau” monetary policy, “allowing its actions to convey all its intended meaning” like the French mime.
But let's get back to transparency about the goal. Both the Fed, Bank of Canada (BoC), and Riksbanken has some sort of dual mandate. They are not only supposed to keep inflation low, but also to keep unemployment low. The Fed has made no official statement as to what level of inflation they aim at. The BoC and Riksbanken, on the other hand, has officially stated that they see two percent inflation as their goals. But none of them has said which inflation figure is supposed to be kept at two percent. Is it CPI or core CPI (or CPIF in Sweden's case)? BoC and Riksbanken are more transparent than the Fed, but they too could improve in this regard.

Another problem is that neither BoC nor Riksbanken has any official goal for the unemployment part or how they weigh inflation compared to unemployment. Is one percent too high inflation equally bad as one percent too high unemployment? Lars Svensson recently proposed that central banks should make all of this official.
Furthermore, an inflation index, a measure of resource utilization, and a measure of stability need to be specified. It is important not to confuse measures of resource utilization to be used as an indicator of inflationary pressure and as a target variable. As the latter, the unemployment gap between unemployment and the sustainable unemployment rate seems to be more reliable and transparent than the alternatives. I am convinced that the framework is more effective if only one inflation index and only one measure of resource utilization is chosen. With multiple inflation and resource-utilization measures, the framework becomes more opaque and accountability becomes difficult to enforce. With many measures, policymakers can often find at least one or two that are close to the desired level and thus motivate quite different policies.
Even if he doesn't go so far as to propose an official exact weighting of inflation versus unemployment, I'll borrow his objective function (from the same paper).

L = (p - p*)^2 + lambda (u - u*)^2,

where p-p* is the deviation of inflation from its target and u-u* the deviation of unemployment. lambda is a positive constant that the central bank should decide. L is the loss function that the central bank will try to minimize. What if the central bank officially adopted this function as their goal? How's that for transparency of the goal? Bernanke could wear a t-shirt with this equation printed on both sides (with p*, u*, and lamda replaced with the official figures).

Well, there's an even better option. How about an official growth path for nominal GDP? That would make the future effects of monetary policy on the economy even more transparent. It would also make the goal itself a bit more transparent since u* can change over time, while the NGDP growth path lasts forever.

2 comments:

  1. Good post Peter.

    We have to be careful when using the word "mandate" for the BoC. It can mean two things:

    1. The 193? Bank of Canada Act, which set up the BoC, and said it had to try to mitigate fluctuations in prices, output and employment. That's a dual (or treble) mandate. But unlike the Fed's mandate, which tells it to go for high (maximum?) employment, the BoC's mandate is much more in tune with New Keynesian thinking, because it is told to reduce fluctuations in employment.

    2. Every 5 years the inflation targeting agreement is jointly renewed by the BoC and government. (It's up for renewal now). This agreement only says 2% inflation, with no mention of employment. (But because the BoC has a 2 year horizon for targeting inflation, it sees no big conflict between targeting inflation and mitigating fluctuations in employment).

    ReplyDelete
  2. Thanks Nick!

    It always gets more complicated the more closely you look at it. The Fed also has some kind of triple mandate and Riksbanken too I believe. It's almost like it's made this way on purpose. A never ending stream of warning signs: Turn back! This is too complicated for you! You will never understand this! Go home! :)

    ReplyDelete