What makes the Underlying Inflation Gauge (UIG) unique is its power to distinguish between changes in the cost of living and changes in the rate of inflation. Did you think those were the same thing? Think again, and read on.
What’s the difference?
The concept of the cost-of-living stems from the first of those role of money as a medium of exchange. When we say the cost of living increases, we mean that it gets harder to maintain a given standard of living on a given income. Either we have to be satisfied with fewer goods or services, or save less, or work harder. In the language of economics, a change in the cost of living is a real phenomenon.
Inflation, in concept, is best understood a change in the value of our unit of account, the dollar. When there is inflation, the value of the unit is smaller each day than it was the day before, for all transactions.
Imagine that you woke up one morning to find that someone had chopped an inch off all our rulers, so that today’s foot was now only as long as yesterday’s eleven inches. You might go from being six feet tall to six-foot-six, but it wouldn’t be any easier for you to reach the top shelf in the kitchen without a footstool. Similarly, if inflation raises both your income and the prices of everything you buy by the same percentage, the value of a dollar as an economic ruler shrinks, but it is neither harder nor easier to maintain the same real standard of living. In that sense, inflation does not measure anything real. It is a purely nominal phenomenon.
There are two other important differences between inflation and changes in the cost of living.
First, although they are both harmful, they are harmful in different ways. An increase in the cost of living hurts people because it makes them poorer. The harm from inflation is more subtle. Inflation makes it harder to plan for the future, so it discourages investment. It erodes the real value of cash and other assets that have fixed nominal values, so it discourages saving. Because the rate of inflation typically becomes more variable as it becomes more rapid, it increases uncertainty. People can overcome some of the uncertainty by indexing the contracts they make, but indexing is costly and never perfect. When we take all these effects together, inflation makes markets work less efficiently and slows economic growth.
Second, the effects of inflation are the same for everyone, but changes in the cost of living vary from place to place and from person to person. If inflation shrinks the unit of account by 3 percent, then the real values of anything with a fixed nominal value — a paycheck, of a Treasury bond, or of a contract to deliver goods — all fall by 3 percent. In contrast, a change in a broad index like the consumer price index (CPI) affects people’s cost of living differently according to which components change. An increase in the price of heating fuel affects people in cold regions more than those in warm regions. An increase in the price of meat does no harm to vegetarians. The cost of living in New York does not necessarily keep pace with that in Oklahoma City.
Can we measure inflation and the cost of living separately?
The CPI is primarily a measure of the cost of living. Each month, the BLS gathers data on the prices of hundreds of goods and services. After making adjustments for changes in package sizes, and, periodically, for changes in quality, it reports the result as a weighted average. If your consumption patterns are reasonably close to the average, the change in the CPI tells you how much harder or easier it is to maintain your standard of living on a given nominal income.
We can take the CPI at face value as a measure of the cost of living, but we have to do some work to extract information from it about the rate of inflation.
A good first step is to adjust the CPI for predictable seasonal changes in the cost of particular goods. For example, everyone knows that the price of fresh vegetables goes up when the summer harvest season is over, but if we expect the price to go down by the same amount next summer, we don’t count the seasonal change as inflation. Consequently, the BLS reports the monthly CPI in both a seasonally adjusted version and a version without seasonal adjustment. Adjusting the raw CPI for predictable seasonal changes greatly smooths month-to-month volatility. We can also get rid of seasonal variations by looking at the CPI for any given month compared to its value 12 months earlier.
Another way to extract an inflation signal from the noise of month-to-month changes in the CPI is to sort out price changes that have particular, microeconomic causes from those that have more general, macroeconomic causes. Microeconomic causes include things like extreme weather events, wars, or developments in global commodity markets. Macroeconomic causes include changes in monetary policy, changes in fiscal policy, and the state of the business cycle.
Food and energy prices are particularly susceptible to transient microeconomic events. To sift them out, the Bureau of Labor Statistics publishes a so-called core CPI, which removes the food and energy components. As the following chart shows, the core CPI is considerably less volatile than the CPI itself, even when the CPI is averaged over 12 months, so it gives us a better idea of the underlying rate of inflation.
An even better way to measure inflation
But, economists at the New York Fed think there is an even better way to measure inflation. Instead of removing data from the CPI to get rid of the noise, as the core CPI does, why not add data to get more information on underlying trends in the unit of account? The additional data used by the New York Fed includes real variables, like employment and output data; monetary data like trends in the money stock; and financial data like interest rates and bank credit. Subjecting all of this data to a complex filtering procedure gives the Underlying Inflation Gauge. The New York Fed supplies monthly values for the UIG going back to 1995.
Here is what we get when we put the three inflation series together:
First, we see that the UIG, like the core CPI, tends to smooth the data. It is not as volatile as the CPI itself, but it is a little more volatile than the core CPI. That already tells us that it has some value in filtering out the inflation signal from the cost-of-living noise.
More importantly, though, the UIG is much more sensitive to cyclical variations in the economy. Although the UIG only goes back far enough to capture two previous recessions, what it shows us is highly suggestive.
In June 2000, the rate of change of the UIG began a sustained decrease some nine months before the start of the short recession that began in March of 2001. The UIG continued to slow throughout the recession. In contrast, the rate of inflation as measured by the CPI remained high and more volatile during the recession, and the rate of increase in the core CPI actually rose.
A similar pattern can be seen going into the much deeper Great Recession that began in December 2007 and lasted until mid-2009. The rate of change of the UIG began to slip well before the recession started, and fell steadily once it was well under way. In contrast, inflation as measured by the CPI itself increased in the early months of the recession, and the rate of change of the core CPI was actually higher a year after the onset of the recession than when it began.
The bottom line
A word of caution: The above is only “eyeball econometrics.” We don’t have a long enough data series for the UIG to give statistically conclusive results. But, with that disclaimer in mind, if you’re looking for early warning signs of an impending recession, I’d keep my eye on the UIG rather than either of the other two measures of price trends.
The UIG has been falling for 10 months now. True, it gave a false alarm with a similar decline starting in mid-2011, but this time the downturn in the UIG coincides with other worrying trends in world trade and financial markets. It just might mean something.
Previously posted at Medium.com