Bullard is a serious critic, but I think that his speech falters on a foundational issue that he barely even addresses: why should the headline CPI be a goal at all? He takes for granted that headline CPI should be the medium-term target of monetary policy, and that targeting the core is just a (flawed) short-term approach for achieving it:
Measures of overall, or headline, inflation attempt to include changes in the prices paid for a wide variety of goods—that is, what households actually have to pay for their daily purchases. This is a sensible notion of precisely what the central bank can and should control over the medium term.
Many discussions of monetary policy, even within the central banking community, discuss movements of subsets of prices instead of the overall or headline measure of price changes. The most famous subset is the “core”—all prices except those relating to food or energy. Core inflation is the measured rate of increase of these prices. Control of core inflation is not the goal of monetary policy, although it sometimes seems to be given the amount of emphasis put on this concept in the U.S.
Admittedly, targeting the headline CPI may be the traditional goal of monetary policy. But it’s not clear why that tradition is valid, or whether there is any cogent rationale for caring about this particular price index. Bullard finds it self-explanatory that stabilizing the price of “what households actually have to pay for their daily purchases” is a “sensible notion” of what central banks should do—yet I could use very similar rhetoric to justify stabilizing what households earn, and implementing a nominal wage target.
Let’s think about why we might want to stabilize inflation in the first place. First there is the direct cost of inflation, which according to most economists comes via three channels:
- Deviating from the Friedman Rule: inflation widens the gap between the yield on bonds (the nominal interest rate) and the yield on money (usually zero), leading consumers to hold an inefficiently low amount of money. As I’ve argued before, the practical relevance of the Friedman Rule is surely minimal. But even if we assume there’s a significant effect, there’s no reason to stabilize inflation today: after all, the nominal interest rate is currently very close to zero, just as the Friedman Rule would suggest. Changes in oil prices today only matter to the extent that they cause an increase in trend inflation several years in the future, when the Fed actually starts raising rates above zero—an increase which, barring some sudden erosion in credibility, seems very unlikely to happen.
- Price dispersion. In most New Keynesian models, the cost of inflation arises from price stickiness. Since many prices are adjusted irregularly, inflation leads to “price dispersion”, as prices that have been adjusted recently are unnaturally high relative to those that have not. But of course, the reason that food and energy prices were excluded from the core CPI in the first place is that they are not sticky—in fact, they change all the time, and they’re extremely volatile. If you’re concerned about price dispersion, you absolutely do not want to target the full headline CPI.
- Interaction with capital gains taxes. This may be the most serious cost of inflation (though sadly, it’s one that is completely avoidable). Since these taxes do not adjust for inflation, investors are penalized for “capital gains” that are really just nominal increases due to inflation. This raises the implicit tax on capital investment. But what kind of “inflation” is relevant here? Since the return to capital is jointly determined by consumers’ preferences and the universe of available investments, we should be looking both at the inflation rate for consumption and the inflation rate for the marginal return to investment. The former might be captured by the headline CPI, but the latter is very different, and it’s probably closer to the GDP deflator.
In short, looking at the direct costs of inflation, I see little to support Bullard’s contention that headline CPI should be our target of choice.
Next there’s the fact that stabilizing inflation is a way to achieve macroeconomic stabilization in general. But as I discussed in a previous post, there is no reason why any form of the CPI should be appropriate for this, and the headline CPI is particularly inapt. Mankiw and Reis’s paper, which uses a formal model to derive a “stability price index”, shows that the index should put low weights on both (A) prices with low signal-to-noise ratios as macroeconomic indicators and (B) prices that are flexible. These are precisely the kinds of prices (food and energy) that are excluded from the core CPI! Since they reflect supply and demand on a world market, they’re very noisy indicators of the American economy, and their prices are adjusted every day.
During the second half of 2008 and into 2009, headline inflation measured from one year earlier fell dramatically and in fact moved into negative territory. This was a signal—one among many, to be sure—that a dramatic shock was impacting the U.S. economy.
It’s very hard for me to see how food and energy prices provided much useful information to the Fed in late 2008 and early 2009—it was easy enough to see equity prices and inflation expectations collapsing. In fact, during the first half of 2008, increases in energy prices arguably misled the Fed into holding back on interest rate cuts when it should have been more aggressive. The decision to keep the federal funds rate at 2% until October was in retrospect a very, very serious mistake, and headline inflation was one of the culprits.
Finally, one of the most commonly cited reasons for stabilizing inflation is the importance of nominal contracts. The idea is that you want the meaning of “one dollar” to be predictable over time, so that when contracts are written in nominal terms there isn’t any avoidable uncertainty. But how should we stabilize the meaning of a dollar—with respect to wages or consumption? As I’ve discussed before, there isn’t any answer that’s obvious from first principles. In some models, where assets are priced according to the marginal utility of consumption in each state of nature (which gives you a “stochastic discount factor“), you’d want to stabilize the price of consumption. But in the real world, there is also a great deal of importance placed on credit constraints—you really, really want to be able to pay your debts, and your ability to pay those debts is closely related to your nominal income. (These concerns are particularly salient during a recession.) This suggests that we should emphasize wages as well. If an oil shock comes along and raises the cost of consumption by 3%—while barely increasing wages at all—it isn’t prudent to try and push all other prices (and therefore wages) 3% below trend to make up the difference.
All in all, I can’t see why Bullard is so confident that the headline CPI is more important than the core. To be sure, core inflation is a clumsy approximation to the ideal price index—we should be incorporating issues like sticky prices and signal-to-noise ratios in a much more systematic and coherent way. But in practice it’s still an improvement, and it wouldn’t have kicked us off track in the catastrophic summer of 2008.