On March 1, the American Institute for Economic Research (AIER) released a fascinating and helpful set of statistics regarding inflation. The study analyzes the popular and widely reported Consumer Price Index (CPI), and explains why it is incorrect. The CPI is published by the Bureau of Labor Statistics (BLS) within the U.S. Department of Labor.
All creators of inflation indexes start with a collection (often called a basket) of items to be purchased. The prices of what are in the basket are accumulated as time passes to show the increase (inflation) or decrease (deflation) that is occurring.
The underlying issue is that what is being purchased changes over time. To determine what should be in the basket, economists creating the price index survey consumers to determine what is being purchased, and in what quantities. AIER provides the following example:
… in the early 1980s, medical studies and a general increased interest in health led people to consume less beef and more chicken and fish. They also became more likely to join health clubs. In addition, as incomes increased, people chose to eat out more and eat at home less. BLS surveys picked up these trends, and the bureau adjusted the CPI accordingly….
Besides weighting, another factor that impacts long-term analysis is the emergence of new products and services. In 1997, the BLS significantly changed the list of goods and services that make up the CPI. Many new products, such as mobile phones, movie discs, and Internet services, were added. Other products were reclassified into new categories. Instead of tracking the cost of a long-distance phone call, for example, the new classification simply tracks the cost of telephone services. This partly reflects the emergence of long-distance telephone plans that no longer charged per minute but rather used a fixed fee.”
AIER has no quarrel with the BLS changing what comprises the CPI basket of purchases. However, they do believe that the government’s basket does not reflect what is actually being purchased. Although the differences/weightings are examined in much greater detail in the AIER study, I have boiled most of the difference to the following:
- The BLS data takes credit for a disproportionate decrease in the cost of decreasing electronic goods.
- Food and beverage costs have been increasing to a greater extent than BLS calculates, and are a large proportion of the typical budget. Because these are large purchases, an even small understating of the weight given to food and beverage costs will have an important understatement of the CPI.
- Fuel and energy costs are rising more than other products and services, and are not given sufficient weight in the CPI.
A prior blog post articulated the challenge with this government statistic. In this prior post, food and energy prices were again the focus.
AIER calls their index the “everyday price index” (EPI). AIER provides the data underlying their conclusions, and encourages anyone interested to calculate a personal inflation index for what that person is actually purchasing. AIER concludes that inflation calculated with what most consumers are actually purchasing will show inflation that is at least double what the U.S. government is reporting. Here is how AIER states it:
In essence, we each have our own Everyday Price Index. People who spend more of their money on products with rapid price increases have seen their cost of living rise even more dramatically than the EPI suggests. This includes people who spend more on fuel and transportation, prescription drugs, tobacco, cable TV (part of the recreation category), and child care.
On the other hand, people who are very healthy or who do not smoke are not at all affected by rapid price increases in prescription drugs and tobacco. People who spend more on products that are experiencing falling prices—such as personal care products and services, household supplies, food and beverages, and phone and Internet service—saw their everyday cost of living rise more slowly than the EPI suggests. But the everyday cost of living is likely to have risen faster that the official CPI would imply for everyone. (Emphasis added)
Prior to 2002, CPI inflation may have been a reasonable approximation for the price increases people faced in their everyday purchases. But this is no longer the case.”
The American Institute for Economic Research is not the only organization stating that the CPI is not adequately measuring the higher inflation that our economy is experiencing. Another blog entry discussed this same issue, and an alternative measure that two economists at MIT calculated. The result is the same; namely, the CPI statistics reported by the Department of Labor reports inflation lower than what is really occurring.
Regardless of these shortcomings cited in the above studies, CPI is and will continue to be used for a variety of government entitlement payments, contractual adjustments, tax rate determinations, and valuation studies. Until a better statistic is universally adopted, CPI’s widespread use makes CPI important because it is effectively “baked in” to other statistics and adjustments involving price changes.