The most commonly cited measure of inflation in the United States is the Consumer Price Index (CPI). The CPI is calculated by government statisticians at the U.S. Bureau of Labor Statistics based on the prices in a fixed basket of goods and services that represents the purchases of the average family of four. In recent years, the statisticians have paid considerable attention to a subtle problem: that the change in the total cost of buying a fixed basket of goods and services over time is conceptually not quite the same as the change in the cost of living, because the cost of living represents how much it costs for a person to feel that his or her consumption provides an equal level of satisfaction or utility.
To understand the distinction, imagine that over the past 10 years, the cost of purchasing a fixed basket of goods increased by 25% and your salary also increased by 25%. Has your personal standard of living held constant? If you do not necessarily purchase an identical fixed basket of goods every year, then an inflation calculation based on the cost of a fixed basket of goods may be a misleading measure of how your cost of living has changed. Two problems arise here: substitution bias and quality/new goods bias.
When the price of a good rises, consumers tend to purchase less of it and to seek out substitutes instead. Conversely, as the price of a good falls, people will tend to purchase more of it. This pattern implies that goods with generally rising prices should tend over time to become less important in the overall basket of goods used to calculate inflation, while goods with falling prices should tend to become more important. Consider, as an example, a rise in the price of peaches by $100 per pound. If consumers were utterly inflexible in their demand for peaches, this would lead to a big rise in the price of food for consumers. Alternatively, imagine that people are utterly indifferent to whether they have peaches or other types of fruit. Now, if peach prices rise, people completely switch to other fruit choices and the average price of food does not change at all. A fixed and unchanging basket of goods assumes that consumers are locked into buying exactly the same goods, regardless of price changes—not a very likely assumption. Thus, substitution bias—the rise in the price of a fixed basket of goods over time—tends to overstate the rise in a consumer’s true cost of living, because it does not take into account that the person can substitute away from goods whose relative prices have risen.
The other major problem in using a fixed basket of goods as the basis for calculating inflation is how to deal with the arrival of improved versions of older goods or altogether new goods. Consider the problem that arises if a cereal is improved by adding 12 essential vitamins and minerals—and also if a box of the cereal costs 5% more. It would clearly be misleading to count the entire resulting higher price as inflation, because the new price is being charged for a product of higher (or at least different) quality. Ideally, one would like to know how much of the higher price is due to the quality change, and how much of it is just a higher price. The Bureau of Labor Statistics, which is responsible for the computation of the Consumer Price Index, must deal with these difficulties in adjusting for quality changes.
Visit this website to view a list of Ford car prices between 1909 and 1927. Consider how these prices compare to today’s models. Is the product today of a different quality?
A new product can be thought of as an extreme improvement in quality—from something that did not exist to something that does. However, the basket of goods that was fixed in the past obviously does not include new goods created since then. The basket of goods and services used in the Consumer Price Index (CPI) is revised and updated over time, and so new products are gradually included. But the process takes some time. For example, room air conditioners were widely sold in the early 1950s, but were not introduced into the basket of goods behind the Consumer Price Index until 1964. The VCR and personal computer were available in the late 1970s and widely sold by the early 1980s, but did not enter the CPI basket of goods until 1987. By 1996, there were more than 40 million cellular phone subscribers in the United States—but cell phones were not yet part of the CPI basket of goods. The parade of inventions has continued, with the CPI inevitably lagging a few years behind.
The arrival of new goods creates problems with respect to the accuracy of measuring inflation. The reason people buy new goods, presumably, is that the new goods offer better value for money than existing goods. Thus, if the price index leaves out new goods, it overlooks one of the ways in which the cost of living is improving. In addition, the price of a new good is often higher when it is first introduced and then declines over time. If the new good is not included in the CPI for some years, until its price is already lower, the CPI may miss counting this price decline altogether. Taking these arguments together, the quality/new goods bias means that the rise in the price of a fixed basket of goods over time tends to overstate the rise in a consumer’s true cost of living, because it does not take into account how improvements in the quality of existing goods or the invention of new goods improves the standard of living. The following Clear It Up feature is a must-read on how the CPI is comprised and calculated.
Note: How do U.S. Government Statisticians Measure the Consumer Price Index?
When the U.S. Bureau of Labor Statistics (BLS) calculates the Consumer Price Index, the first task is to decide on a basket of goods that is representative of the purchases of the average household. This is done by using the Consumer Expenditure Survey, a national survey of about 7,000 households, which provides detailed information on spending habits. Consumer expenditures are broken up into eight major groups, shown below, which in turn are broken up into more than 200 individual item categories. The BLS currently uses 1982–1984 as the base period.
For each of the 200 individual expenditure items, the BLS chooses several hundred very specific examples of that item and looks at the prices of those examples. So, in figuring out the “breakfast cereal” item under the overall category of “foods and beverages,” the BLS picks several hundred examples of breakfast cereal. One example might be the price of a 24-oz. box of a particular brand of cereal sold at a particular store. The specific products and sizes and stores chosen are statistically selected to reflect what people buy and where they shop. The basket of goods in the Consumer Price Index thus consists of about 80,000 products; that is, several hundred specific products in over 200 broad-item categories. About one-quarter of these 80,000 specific products are rotated out of the sample each year, and replaced with a different set of products.
The next step is to collect data on prices. Data collectors visit or call about 23,000 stores in 87 urban areas all over the United States every month to collect prices on these 80,000 specific products. A survey of 50,000 landlords or tenants is also carried out to collect information about rents.
The Consumer Price Index is then calculated by taking the 80,000 prices of individual products and combining them, using weights (as shown in Figure 1) determined by the quantities of these products that people buy and allowing for factors like substitution between goods and quality improvements, into price indices for the 200 or so overall items. Then, the price indices for the 200 items are combined into an overall Consumer Price Index. According the Consumer Price Index website, there are eight categories used by data collectors:
The Eight Major Categories in the Consumer Price Index
- Food and beverages (breakfast cereal, milk, coffee, chicken, wine, full-service meals, and snacks)
- Housing (renter’s cost of housing, homeowner’s cost of housing, fuel oil, bedroom furniture)
- Apparel (men’s shirts and sweaters, women’s dresses, jewelry)
- Transportation (new vehicles, airline fares, gasoline, motor vehicle insurance)
- Medical care (prescription drugs and medical supplies, physicians’ services, eyeglasses and eye care, hospital services)
- Recreation (televisions, cable television, pets and pet products, sports equipment, admissions)
- Education and communication (college tuition, postage, telephone services, computer software and accessories)
- Other goods and services (tobacco and smoking products, haircuts and other personal services, funeral expenses)
Figure 1: Of the eight categories used to generate the Consumer Price Index, housing is the highest at 41%. The next highest category, transportation at 16.8%, is less than half the size of housing. Other goods and services, and apparel, are the lowest at 3.4% and 3.6%, respectively. (Source: www.bls.gov/cpi)
The CPI and Core Inflation Index
Imagine if you were driving a company truck across the country- you probably would care about things like the prices of available roadside food and motel rooms as well as the truck’s operating condition. However, the manager of the firm might have different priorities. He would care mostly about the truck’s on-time performance and much less so about the food you were eating and the places you were staying. In other words, the company manager would be paying attention to the production of the firm, while ignoring transitory elements that impacted you, but did not affect the company’s bottom line.
In a sense, a similar situation occurs with regard to measures of inflation. As we’ve learned, CPI measures prices as they affect everyday household spending. Well, a core inflation index is typically calculated by taking the CPI and excluding volatile economic variables. In this way, economists have a better sense of the underlying trends in prices that affect the cost of living.
Examples of excluded variables include energy and food prices, which can jump around from month to month because of the weather. According to an article by Kent Bernhard, during Hurricane Katrina in 2005, a key supply point for the nation’s gasoline was nearly knocked out. Gas prices quickly shot up across the nation, in some places up to 40 cents a gallon in one day. This was not the cause of an economic policy but rather a short-lived event until the pumps were restored in the region. In this case, the CPI that month would register the change as a cost of living event to households, but the core inflation index would remain unchanged. As a result, the Federal Reserve’s decisions on interest rates would not be influenced. Similarly, droughts can cause world-wide spikes in food prices that, if temporary, do not affect the nation’s economic capability.
As former Chairman of the Federal Reserve Ben Bernanke noted in 1999 about the core inflation index, “It provide(s) a better guide to monetary policy than the other indices, since it measures the more persistent underlying inflation rather than transitory influences on the price level.” Bernanke also noted that it helps communicate that every inflationary shock need not be responded to by the Federal Reserve since some price changes are transitory and not part of a structural change in the economy.
In sum, both the CPI and the core inflation index are important, but serve different audiences. The CPI helps households understand their overall cost of living from month to month, while the core inflation index is a preferred gauge from which to make important government policy changes.
Practical Solutions for the Substitution and the Quality/New Goods Biases
By the early 2000s, the Bureau of Labor Statistics was using alternative mathematical methods for calculating the Consumer Price Index, more complicated than just adding up the cost of a fixed basket of goods, to allow for some substitution between goods. It was also updating the basket of goods behind the CPI more frequently, so that new and improved goods could be included more rapidly. For certain products, the BLS was carrying out studies to try to measure the quality improvement. For example, with computers, an economic study can try to adjust for changes in speed, memory, screen size, and other characteristics of the product, and then calculate the change in price after these product changes are taken into account. But these adjustments are inevitably imperfect, and exactly how to make these adjustments is often a source of controversy among professional economists.
By the early 2000s, the substitution bias and quality/new goods bias had been somewhat reduced, so that since then the rise in the CPI probably overstates the true rise in inflation by only about 0.5% per year. Over one or a few years, this is not much; over a period of a decade or two, even half of a percent per year compounds to a more significant amount. In addition, the CPI tracks prices from physical locations, and not at online sites like Amazon, where prices can be lower.
When measuring inflation (and other economic statistics, too), a tradeoff arises between simplicity and interpretation. If the inflation rate is calculated with a basket of goods that is fixed and unchanging, then the calculation of an inflation rate is straightforward, but the problems of substitution bias and quality/new goods bias will arise. However, when the basket of goods is allowed to shift and evolve to reflect substitution toward lower relative prices, quality improvements, and new goods, the technical details of calculating the inflation rate grow more complex.
Additional Price Indices: PPI, GDP Deflator, and More
The basket of goods behind the Consumer Price Index represents an average hypothetical U.S. household, which is to say that it does not exactly capture anyone’s personal experience. When the task is to calculate an average level of inflation, this approach works fine. What if, however, you are concerned about inflation experienced by a certain group, like the elderly, or the poor, or single-parent families with children, or Hispanic-Americans? In specific situations, a price index based on the buying power of the average consumer may not feel quite right.
This problem has a straightforward solution. If the Consumer Price Index does not serve the desired purpose, then invent another index, based on a basket of goods appropriate for the group of interest. Indeed, the Bureau of Labor Statistics publishes a number of experimental price indices: some for particular groups like the elderly or the poor, some for different geographic areas, and some for certain broad categories of goods like food or housing.
The BLS also calculates several price indices that are not based on baskets of consumer goods. For example, the Producer Price Index (PPI) is based on prices paid for supplies and inputs by producers of goods and services. It can be broken down into price indices for different industries, commodities, and stages of processing (like finished goods, intermediate goods, crude materials for further processing, and so on). There is an International Price Index based on the prices of merchandise that is exported or imported. An Employment Cost Index measures wage inflation in the labor market. The GDP deflator, measured by the Bureau of Economic Analysis, is a price index that includes all the components of GDP (that is, consumption plus investment plus government plus exports minus imports). Unlike the CPI, its baskets are not fixed but re-calculate what that year’s GDP would have been worth using the base-year’s prices. MIT's Billion Prices Project is a more recent alternative attempt to measure prices: data are collected online from retailers and then composed into an index that is compared to the CPI (Source: http://bpp.mit.edu/usa/).
What’s the best measure of inflation? If concerned with the most accurate measure of inflation, use the GDP deflator as it picks up the prices of goods and services produced. However, it is not a good measure of cost of living as it includes prices of many products not purchased by households (for example, aircraft, fire engines, factory buildings, office complexes, and bulldozers). If one wants the most accurate measure of inflation as it impacts households, use the CPI, as it only picks up prices of products purchased by households. That is why the CPI is sometimes referred to as the cost-of-living index. As the Bureau of Labor Statistics states on its website: “The ‘best’ measure of inflation for a given application depends on the intended use of the data.”
Key Concepts and Summary
Measuring price levels with a fixed basket of goods will always have two problems: the substitution bias, by which a fixed basket of goods does not allow for buying more of what is relatively less expensive and less of what is relatively more expensive; and the quality/new goods bias, by which a fixed basket cannot take into account improvements in quality and the advent of new goods. These problems can be reduced in degree—for example, by allowing the basket of goods to evolve over time—but they cannot be totally eliminated. The most commonly cited measure of inflation is the Consumer Price Index (CPI), which is based on a basket of goods representing what the typical consumer buys. The Core Inflation Index further breaks down the CPI by excluding volatile economic variables. Several price indices are not based on baskets of consumer goods. The GDP deflator is based on all the components of GDP. The Producer Price Index is based on prices of supplies and inputs bought by producers of goods and services. An Employment Cost Index measures wage inflation in the labor market. An International Price Index is based on the prices of merchandise that is exported or imported.
Bernhard, Kent. “Pump Prices Jump Across U.S. after Katrina.” NBC News, September 1, 2005. http://www.nbcnews.com/id/9146363/ns.../#.U00kRfk7um4.
Wynne, Mark A. “Core Inflation, A Review of Some Conceptual Issues.” Federal Reserve Bank of St. Louis. p. 209. Accessed April 14, 2014. http://research.stlouisfed.org/publi...art2/Wynne.pdf
- Consumer Price Index (CPI)
- a measure of inflation calculated by U.S. government statisticians based on the price level from a fixed basket of goods and services that represents the purchases of the average consumer
- core inflation index
- a measure of inflation typically calculated by taking the CPI and excluding volatile economic variables such as food and energy prices to better measure the underlying and persistent trend in long-term prices
- Employment Cost Index
- a measure of inflation based on wages paid in the labor market
- GDP deflator
- a measure of inflation based on the prices of all the components of GDP
- International Price Index
- a measure of inflation based on the prices of merchandise that is exported or imported
- Producer Price Index (PPI)
- a measure of inflation based on prices paid for supplies and inputs by producers of goods and services
- quality/new goods bias
- inflation calculated using a fixed basket of goods over time tends to overstate the true rise in cost of living, because it does not take into account improvements in the quality of existing goods or the invention of new goods
- substitution bias
- an inflation rate calculated using a fixed basket of goods over time tends to overstate the true rise in the cost of living, because it does not take into account that the person can substitute away from goods whose prices rise by a lot