Chained CPI: What it is and Why You Should Care

Chained CPI: What it is and Why You Should Care

In recent years, the term “chained-CPI” has crept into news stories regarding the federal budget and the solvency of Social Security. Many people are familiar with the term because of the news stories and political debates. But many don’t really understand what a chained-CPI is, and therefore don’t really understand what the controversy is all about. I’ll try to explain it in everyday language.

A chained-CPI is a price index which has been proposed as a method for cutting federal spending by reducing future monthly benefit payments to retirees, disabled persons, and veterans. This includes all Social Security payments. Sometimes veterans’ benefits are left out of the discussions, and the focus is on trying to “fix” a broken Social Security system.

By switching to a chained-CPI, future cost of living adjustments (COLAs), intended to prevent people on a fixed income from losing purchasing power due to inflation, would be lower than under current procedures. It is important to note that when you hear talk of benefit cuts, the proposals involving a chained-CPI wouldn’t actually cut current benefit payments – only future cost of living increases in payments would be cut.

In order to understand how this would work, and why it is a bad idea, first we need to understand some important points regarding:

  1. Realities of inflation
  2. Methods for measuring inflation
  3. Flaws inherent in measures of inflation

Then, we can see how a chained-CPI measurement differs from the current method for calculating the annual COLA, and why it would be a bad idea for determining benefits for the specific groups of people being targeted.

Realities of Inflation

The U.S. economy is large and complex, and it includes – literally – millions of different prices. In a healthy market-based economy, individual prices are constantly changing. Millions of prices, constantly being updated due to market forces, do not all change at the same rate, at the same time, or even always in the same direction. Some prices go up, some go down. Some changes are rapid, some are slow. Some prices are volatile, some are relatively stable. If you want a visual example, take a look at the daily listing of stock market prices. You will likely see a headline saying what the change in the index is for the day. On average, this number will be a positive number because stock prices tend to increase in the long run. But take a look at the individual prices listed. On a normal day, you will see many with an uptick, and many with a downtick. Some prices will have no change for the day. Some of these ticks are much larger than others. Individually, the price changes do not match the average. This is how stock prices work. It is also how price changes work for the goods and services we buy.

Wage rates are prices. One person’s income is somebody else’s expense. That is another reality in a market-based economy.

The general trend over time is for prices (or prices and wages, depending on the terminology you prefer) to go up rather than stay the same or go down. This is a general trend, an average of millions of prices which individually do not follow the same trend.

The point I am trying to make is that even though we are told that the rate of inflation is a certain percentage, that rate is only a measurement based on an average of different price changes which do not match this average. As a result, inflation will affect different people in different ways, depending on their individual spending habits and other factors.


Methods for Measuring Inflation

During any specific time frame, inflation will affect different people in different ways. Price indexes have been developed as an attempt to measure the average effect of inflation on different groups of people. Various types of price indexes are available, but the ones which are relevant to this discussion are Consumer Price Index (CPI) measurements. Specifically, the U.S. Bureau of Labor Statistics (BLS) publishes various CPI measurements which are relevant to this discussion.

Given that the economy is large and complex, these measurements are estimates. They are attempts at measuring average, not actual, effects of inflation on individuals. The CPI measurements are based on price changes in a predetermined list of goods which has been designated as a “typical” bundle of goods purchased by a “typical” household in a given period of time.

The CPI measurements we need to focus on for this discussion are the CPI-U and the CPI-W, as well as the Chained Consumer Price index (C-CPI-U).

When you see a news report which mentions what the rate of inflation is for a specific time frame, the measurement being used is usually the CPI-U. The “U” stands for “All Urban Consumers”. This is the measurement that is typically used whenever you see a report such as “a $4,000 car in 1965 would be worth $30,150.22 in today’s dollars”. In case you want to calculate the effects of inflation on a specific dollar amount between two different time periods, here is a link to the BLS Inflation Calculator.

The CPI-W is similar to the CPI-U, except that instead of “All Urban Consumers”, this one attempts to measure the effects of inflation on “Urban Wage Earners and Clerical Workers”. This measurement is most often used in labor contract negotiations, since its specific focus is on wage earners.

The chained CPI (C-CPI-U), which is the main focus of this essay, is an alternative measurement based on a known flaw in the method for calculating CPI-U and CPI-W.

Flaws inherent in measures of inflation

There are a number of reasons why any single measurement of inflation will be a flawed measurement. I’ll get to a list of several of these reasons, but first I want to focus on the one which separates the chained CPI from other CPI measurements. This will not only explain how a chained-CPI is calculated, but it will also explain the theoretical justification for using this alternative method.

I already mentioned that CPI-U and CPI-W are based on a predetermined “typical” bundle of goods purchased by a “typical” household within a specific time frame. This “typical” bundle of goods is not only predetermined, it remains the same throughout the time frame. In other words, it is a fixed set of goods. The same items and the same quantities of each item go into the calculations throughout the time frame. One flaw in using a fixed set of goods is that doing so does not factor in any substitution effect. Remember from earlier in this discussion that inflation is an average of changes in different prices which move at different rates and perhaps in different directions, with a general trend of higher prices. When this happens, a fixed amount of income will typically be used to purchase fewer items with large price increases, and more items with either price decreases or smaller price increases. In other words, when the relative prices of goods within a bundle of goods change, then the bundle of goods no longer is fixed. A substitution effect takes place. The same consumer will change his spending pattern over time based solely on relative price changes.

The chained CPI is simply a CPI-U measurement which factors in this substitution effect. Instead of a “fixed” bundle of goods, it uses a “chained” bundle of goods. Theoretically, this adjustment for the substitution effect makes the measurement more in line with the price differences that consumers actually pay, not the price differences that the consumer would have paid under a fixed bundle of goods.

Using a fixed bundle of goods for the calculations is thought to place an upward bias on the resulting index of inflation. The justification for switching to a chained-CPI is that it corrects for a known flaw in the current method of measuring inflation. It is important to note that this adjustment is always away from higher price increases and towards price decreases and lower price increases. This means that using the chained CPI will result in a smaller measurement for inflation, and smaller cost of living increases for people on a fixed income.

That justification would be valid if the substitution effect were the only, or even the most important, flaw in CPI measurements. Switching to a chained CPI would be justified if it were true that under the current system, senior citizens living on Social Security were increasing their purchasing power year after year because their incomes are based on a higher rate of inflation than their actual increases in living expenses. Does anybody really believe that to be the case? Is increased purchasing power for retirees, disabled persons, and veterans really a problem in the United States? Are we allowing our senior citizens to get richer because of the way we calculate their fixed incomes?

Of course that isn’t the case. Yet that is precisely the economic justification for switching to a chained CPI measurement. In truth, the CPI measurements – all of them, including the chained CPI – are flawed for a number of other reasons which, cumulatively, more than offset the upward bias due to the substitution effect.

  1. In reality, there is no such thing as a “typical” bundle of goods purchased by a “typical” household.
  2. The concept of using a representative sample of goods for measuring inflation means that the resulting inflation index is nothing more than an estimate of an average.
  3. This average is designed to mirror the spending patterns of “all urban consumers” or “urban wage earners or clerical workers”, not the spending patterns of those who receive Social Security benefits or otherwise live on a fixed income.
  4. People on a fixed income tend to have less flexibility in terms of substitution than “all urban consumers” or “urban wage earners or clerical workers”. Senior citizens tend to spend more than others on necessities with high rates of price increases – such as medication, healthcare, and home heating – and less on nonessential items from decreasing-cost industries, such as electronics.
  5. COLAs are based on last year’s inflation rate, meaning that people on fixed incomes are always a year behind in having their incomes adjusted for increases in their living expense.

  6. Measurements of inflation have other known flaws that I didn’t mention above. Core inflation vs. headline inflation is one issue; technology and differences in quality create other problems with measuring inflation. None of these flaws make the case for lowering benefits paid to retirees and veterans through the use of a chained CPI measurement.

    The bottom line is that COLAs do not overstate the true cost of living increases faced by Social Security recipients and veterans. We know this to be true because people living on a fixed income are not increasing their purchasing power over time. Instead, they are having more and more problems making ends meet. Lowering COLAs through a chained CPI would only make this situation worse.

    Politicians are trying to balance the budget on the backs of our most vulnerable citizens.
    Politicians are trying to make senior citizens pay for political mishandling of the Social Security system, when much better options are available.
    Seniors and veterans should be honored, not thrown under a bus.

    A version of this essay is included as a chapter in the book Sanity and Public Policy: Separating Truth from Truisms by Jerry Wyant. This book is available in both paperback and eBook formats.

    Paperback version from Amazon
    Kindle version from Amazon
    All eBook formats from Smashwords

Jerry Wyant