There is no such thing as inflation
In 1987, UK Prime Minister Margaret Thatcher whipped up a firestorm of criticism from her opponents on the left when she told a magazine reporter that “there is no such thing as society”, only individual men and women, and families.
The interpretation of those comments remains fiercely controversial. From the context it is not certain the prime minister was clear what she was trying to say.
But according to one interpretation the prime minister was encouraging her listeners to look beyond the impersonal aggregate of “society” to the individuals behind it.
The distinction between aggregates and individual components is something the Federal Reserve should bear in mind as officials mull whether to launch a new round of asset purchases to keep inflation from falling further and stimulate the recovery.
Because in some sense there is no such thing as inflation, only a collection of price rises for individual items, some rising faster and some slower.
It is clear price increases do have a structural component. Policymakers and economists distinguish between a general rise in the level of prices (“inflation”) and relative price increases for individual items (Adam Smith’s “invisible hand” guiding the reallocation of scarce resources).
But in an economy characterized by uneven spare capacity, with bottlenecks in some areas and unused capacity in others, excess demand and inflationary pressures may not show up evenly. Even as all prices rise (inflation), price rises are likely to be largest in those parts of the system with the worst bottlenecks, while increases in areas suffering significant under-employment of resources lag behind.
The problem for the Fed is that the prices of some items that are in relatively short supply internationally (iron ore, oil, foodstuffs, cotton) are rising quickly while other items where there is a local surplus (U.S. housing, U.S. manufacturing and U.S. wages) are rising more slowly. There is no guarantee adding more liquidity to the economy will close the gap or even affect all prices equally.
By the time the Fed has added enough liquidity to overcome structural problems and stimulate a modest rise in U.S. house prices, manufacturing prices and wages, it will probably have caused a conflagration in oil, metals, grains and other internationally traded commodities. Monetary policy is far too blunt a tool for encouraging the type of targeted and carefully calibrated price increases that Fed officials want.
NOT ONE RATE BUT MANY
Focusing on the “average” rate of price increases is apt to mislead. There is not one single rate of inflation but a whole family of them:
(1) In the United States, policymakers have at different times identified “inflation” with the consumer price index for all urban consumers (CPI-U); the consumer price index excluding food and energy items (core CPI-U); the GDP deflator for personal consumption expenditures (PCE); and the market-based personal expenditures deflator excluding food and energy. The inflation rates given by these measures can and have been utterly different. They don’t even necessarily move in the same direction.
(2) In the United Kingdom, inflation rates given by the retail prices index (RPI) and consumer price index (CPI) have been so different in recent years that the government is shifting indexation of at least some parts of the budget from the (higher) RPI to the (lower) CPI in a bid to cut costs.
(3) In response to popular and media complaints that published inflation rates are not consistent with households’ own experience of rising prices, and generally understate the rate of increase, statisticians across North America and Europe have pointed out that:
(a) Published inflation rates are averages — individual households may have higher or lower personal rates depending on the basket of goods and services they buy.
(b) Personal impressions are skewed when prices for low-value frequently bought items (food and fuel) are rising rapidly while more expensive but less frequently purchased items (consumer electronics and durables) are rising more slowly or even falling.
Inflation rates and price indexes are useful short-hand ways to think about the rate of change in a broad basket of prices. But they are a very imperfect reflection of the “cost of living” let alone trying to fine tune the economy. Every person reading this article will have their own experience of “inflation”.
The problem for the Fed is that different components of the consumer price index are diverging and there is no easy way to stimulate increases in housing, manufactured items and wages without triggering an even more rapid escalation in the cost of energy and food (which could cut spending further since these are costs for most households and corporations).
Charts 1 and 2 show the 12-month increase in the three major components of the U.S. consumer price index — food and energy; shelter (owners’ equivalent rent); and other items. Food/energy and shelter each account for about a quarter of the total weighting, other items account for just under half.
Food and energy have been much more volatile that the rest of the index and have become increasingly unstable in recent years. Inflation in shelter and other items has been much more stable and gently trending downwards for the last two decades.
The Fed’s problem is that while food and energy prices are still rising briskly (2.16 percent in the twelve months to September), other items barely increased (up 0.47 percent) and the cost of shelter as measured by owners’ equivalent rent actually fell (down -0.10 percent).
This is not the first time inflation in the “other items” category has fallen very low. But in 2003 and again in 2005-2006, ultra-low inflation for other items was offset by faster increases in shelter and food/energy. This time shelter costs are falling and food/energy costs are going up only relatively slowly.
The pattern of price rises looks structural rather than cyclical. Inflation in shelter and other items has been on a downtrend for 20 years due to globalisation, off-shoring, better supply chain management, and better wage control, all of which are due to secular trends rather than a short term loss of demand. The sudden disappearance of inflation in the shelter sector is the legacy of the housing bust. Much of the current low inflation problem looks like a simple after-effect of the housing crash.
Both the shelter and the other items components of the CPI have been remarkably stable over the last two decades. To generate an appreciable increase in these two components of the CPI which account for 25 percent and 50 percent of the index respectively, the Fed would probably need to apply an enormous amount of stimulus, buying hundreds of billions if not trillions of financial assets.
The food/energy component in contrast is much more unstable. Because these markets have much less spare capacity, inflation rates have remained higher. If the Fed applied enough stimulus to lift shelter and other items appreciably, it would probably cause a huge increase in the food/energy component as well as an enormous bubble in asset prices such as bonds and equities.
The same problem is being replicated within the energy/food complex itself. Prospective quantitative easing is already causing prices for commodities thought to be in short supply (gold, grains, copper and cotton) to rise much faster than those with spare capacity and slack fundamentals (crude oil and natural gas).
Before they launch a massive asset purchase programme, Fed officials need to ask precisely what sort of inflation they are trying to stimulate, and how QE will achieve it without triggering lots of unintended and harmful side effects.