For many years now, the Consumer Price Index (CPI) has been used as the standard gauge for inflation in different economies. It measures the changes in the cost of a fixed basket of products and services consumed by ordinary consumers in a specific period of time (Moulton 4). To most economists, inflation is caused by a general increase of money in a specific economy.
In this case, the volumes of money being circulated exceeds the supply of goods and services in the same market thus leading to an upward adjustment of prices in order to absorb the extra monies in the economy. But why is measuring inflation important in economies, one may ask?
Well, In the United States for example, the government factors in inflation when paying its citizens for programs such as social security and retirement benefits. This means that erroneous representation of inflation rates would lead to erroneous government financial projections.
This research paper has established that the CPI as used to measure inflation has several biases. As such, the study seeks to pinpoint the biases.
Biases as identified by the Boskin commission
In 1995, the US senate appointed the Boskin Commission to study the computation of CPI and reveal any biases that could exist therein. On December 1996, the commission issued a report which declared that several biases had led to an overstatement of inflation by 1.1 percent in 1996 alone and 1.3 percent in prior years (Gordon 4). Baker supports the Boskin Commission Report’s findings by stating that there is an approximate 0.5 %- 1.5% overestimation of inflation annually in the United States (1).
While explaining the biases that exist in the CPI, the report stated that the major cause of bias was the use of fixed weight index as a representation of the cost of living (Boskin Commission Report VI). According to the report, the cost-of-living Index should be ideally used as a reference point and should be attained using other research methods rather than through the fixed-weight index.
The report further states that, “the fixed-weight index exaggerates the effect of price changes on the cost of living, because it fails to allow for substitutions that enable consumers to avoid the full impact” (VI). This is especially so because as people become wealthier, they demand more products and services, to which the market responds by supplying more dynamic and diverse products and services. In such a situation, it becomes inherently challenging to use the fixed-weight index as a suitable measure for CPI.
Substitution Bias
According to QuickMBA, when prices of specific products rise substantially, consumers usually purchase a substitute cheaper product especially when their income levels remain fixed. The fixed-weight Index used to measure CPI however is not flexible enough to capture such substitution, and hence fails to accurately estimate the effect of price increases on consumer’s budgets.
Quality Bias
As technology advances, its use in the production of good and services improves and leads to improved production quality. As a result, services and products are more useful and last longer than was the case in the past. While the Boskin Commission Report (V) acknowledges that the CPI has been able to capture the effect brought about by quality changes in some products, it argues that a lot other products have been ignored by the Index.
This has eventually led to an increased quality-consideration bias across product categories. According to Schulkin, quality bias is often difficult to quantify especially in a market where competition lead most manufacturers to keep improving the quality of their products and services (1).
New product Bias
According to QuickMBA , the Index waits until new products become common in the market for them to be integrated in the Index (2). This in turn means that price reductions associated with the introduction of new products and oftentimes new technology is not captured or reflected in the CPI.
The Boskin Commission Report estimates that the bias occurring in the CPI by not factoring in new products introduced in the market sooner, amounts to an average of 0.25 percent bias annually (V). Notably, most new products are factored in the CPI almost a decade after they penetrate the market. By this time, their prices are markedly lower than the initial introductory prices they retailed in when they were first introduced into the market (Gordon 14)
The Boskin Commission report notes that on introduction of most new products, the prices are usually high, and the products not as good as they evolve later on in the product-development cycle. As the product matures overtime however, production increases, improvements on the product are made and eventually the product retails at reduced prices thus making a larger percentage of the population to afford the same. This however does not excuse the delayed incorporation of new products in
Outlet bias
According to QuickMBA, consumers usually shift from one purchase outlet to another based on the benefits earned from the outlets (4). With the advent of online retailers, more consumers are for instance opting to do their shopping online since they can earn points or discounts. Others have learned the benefits of buying their supplies in bulk and therefore have shifted from purchasing products from retailers and instead have joined wholesale clubs.
Schulkin observes that the availability of generic products which retail at lower prices, the presence of low price stores, and the increased use of mail orders as a purchase method where consumers do not have to pay sales tax, have increased the outlet options for the buyers (1).Usually, the CPI does not consider such changes in purchase behavior, which occur when the consumer have many retail options offering different prices. As a result, the compiled index becomes a biased representation of household spending across the country.
Others
Schulkin also observes that in addition to the four main biases identified in the Boskin Commission report, using the Fixed-Weight Index to measure inflation ignores the many discounts availed to consumers in a given economy (2).
Examples include cash rebates offered when people use their credit cards, promotional prices offered for airline tickets during the low-travel season, promotional prices for products and services and the countless coupons offered to the consumer market by different manufacturers trying to promote their products or services in the market at any given time.
Notably, the CPI is calculated by factoring in the after-tax prices of products and services in an economy. However, it fails to consider the benefits accrued by the taxes. As Schulkin notes, tax proceeds are mainly used by the government for spending in law enforcement, welfare, infrastructure development, deficit reduction and spending in education among other social issues (1).
Taxes imposed on tobacco and alcohol on the other hand leads to reduced consumption of the same by the general population. In the past, CPI completely disregards these benefits. Of late however, Williams reports that the government is understating CPI by a rough 7 percent per year (2). He says that this happens because the government has redefined how CPI is calculated, most notably the introduction of the variable-basket goods that factored in substitution
Gordon Also states that although changes in CPI are published on a monthly basis in the United States, they cannot be revised even when the estimates there in are questionable (11). This in turn means that the CPI is methodologically inconsistent and contains biases which occur because the index cannot be revised backwards.
In a suggestion borrowed by the commissioned mandated to investigate biases in CPI in the US in 1995, Gordon recommends that since the CPI cannot be revised backwards, it should be supplemented with a research-based index, which should be published annually and, which should be more flexible to allow for continued revisions in order to slot in new research findings (18).
Research findings
Errors in CPI measurement often results in faulty payments especially by government and private sector players who rely on the index to adjust the benefits paid to beneficiaries of government programs or workers. An upward CPI bias often raises the benefits paid to the recipients in an equal proportion to the proportion increase of the index annually (Duggan et al. 4).
This means that if CPI was at 1.1 percent in 1998 and 2.1 percent in 1999, benefits advanced to recipients would be proportionate to the percentage difference between the two years. In an upward bias, the recipients usually benefit from an erroneous compensation which is as a result of the misrepresentation of the CPI.
Continued overestimation of CPI is also likely to continue having a ripple effect every year, thus meaning that the economy will continue to have a series of overestimated inflation estimations (Moulton 2; Schulkin 2).
This is especially because economic analysts use CPI in construction of national income accounts thus meaning that biases often lead to misrepresentation of growth and productivity in the economy. An over-estimation of inflation leads to an overestimation of poverty thresholds in any economy, which in turn means that the government underestimates the economic growth and the well-being of the populace.
Baker also found out that though it’s widely thought that most services and products improve in quality as competition among the different providers intensify; this is not always the case (4). An example in the medical insurance sector suggests that most consumers spend significant amount of resources both in terms of money and times shopping around for the most ideal policy to adopt.
Even after making their minds, they are forced to spend some more time filling the forms and should they need to make claims, even more time is spent filling the claim forms and pursuing payments from the insurance firms. Since the CPI just examines health care expenditure based on the monetary value projections, the costs of time wasted comparing the policy providers, and the cost of time lost pursuing claims is ignored.
This study found out since population census are held every ten years in countries like the United States, there is a probability that the population-expenditure weights are not a true reflection of the inflation. Baker notes that in rapid population growth areas, price increases are likely to be higher than slow population growth areas (4).
With population estimates captured ten years apart, it would be challenging to capture population growth rates accurately hence presenting the probability that CPI is undercounted or over-estimated in different regions depending on the pace of the population growth rate.
The Bolskin report recommended several approaches to resolving the biases. Among them was the adoption of a weighted geometric mode of indexing, which would allow the CPI to factor in substitution of products and services by consumers. A superlative index used at the stratum level was also recommended for use in CPI formulation since such would eliminate irrelevant market baskets and speed up the introduction of new products in the basket for indexing (Gordon 18).
In the United States, it is estimated that a third of all federal spending is indexed according to the CPI. Income tax is also indexed according to changes in the CPI thus affecting federal revenues (Baker 1). With the biases discussed above remaining in place deficits in federal spending and federal revenues are both affected.
The 1996 Boskin Commission Report argued that “if the CPI overstated the change in the cost of living by an average of 1.1 percentage points per year over the next decade, this bias would contribute about $148 billion to the deficit in 2006 and $691 billion to the national debt by then” (6). This would make the bias the fourth most expensive undertaking after the country’s social security program, health care program and defense.
With an overestimated CPI, one can therefore conclude that the average growth in wages have been better than predicted. This however does not scrap away wage inequalities which lead to worse living standards for a section of low-income earners in the society. As Baker suggests, this should stress the importance for the government to lay more emphasis on addressing inequality. This would help the overall growth of the economy since the low income earners would be empowered economically by providing them with more disposable income.
An overestimated CPI would also mean that the concerns that future generations would have to put up with declined living standards is unwarranted. As Baker notes, if it is indeed that the mode of indexing has misled economic analysts to believe that the economic growth and the general savings was well below what it should have been, a change in this would mean that indeed the country’s growth was and still is at a healthy level.
Conclusion
While the Laspeyres index, which is based on a fixed-based mode to calculate the CPI has been common use in the past, economists now are suggesting the uses of other indexes such as Paasche, Fisher-Ideal and the Walsh-price as alternatives. Considering the fast pace of changes in contemporary economies however, it is almost obvious that whichever index is used, arriving at an all encompassing CPI will be a challenging undertaking for the government.
Overall however, the government should avoid using any indexes that would either overestimate or underestimate inflation in the country. This then means that a lot of effort needs to be dedicated towards developing a formula that will be used to gauge inflation as accurately as possible.
Works Cited
Baker, Dean. “Revising the Consumer Price Index: Correcting bias, or biased corrections?” Economic Policy Institute Briefing Paper (1998):1-8.
Duggan, James, Gillingham, Robert & Greenlees, John. “Housing Bias in the CPI and its effects on the Budget Deficit and the Social Security trust fund.” Research paper 9701 (1999): 1-19.
Gordon, Robert. The Boskin Commission Report and Its Aftermath. National Bureau of Economic Research. (1999):1-47.
Moulton, Brent. Bias in the Consumer Price Index: What is the Evidence? US department of labor: Bureau of labor statistics 294.1(1996):1-35.
QuickMBA. Consumer Price Index (CPI). Feb. 2010. 5 July 2010. http://www.quickmba.com/econ/macro/cpi/
Schulkin, Peter. “Upward Bias in the CPI.” Challenge Journal 36.1 (1993): 1-7.
The Boskin Commission Report. Towards a More accurate measure of the cost of living. Reports & Studies. Dec. 1998. 5 July 2010. https://www.ssa.gov/history/reports/boskinrpt.html#fn2
Williams, John. Consumer Price Index-“Government Economic reports: Things you’ve suspected but were afraid to ask!” Shadow Government Statistics. Oct. 2004. 5 July 2010. http://www.shadowstats.com/article/consumer_price_index