BIAS IN CPI

Bias in CPI

           

Consumer Price index seems to be overstating the cost of living by 0.5 to 1.5 percent a year. This has happened since the beginnings of using this formulation but Bias only appeared recently.

Senate Finance Committee has been investigating this Bias and stated than in the recent years CPI has been overstated up to 1.5 %. This has an impact on the well being of the society as it could have dramatic price increases in the long term.

Bureau of labor statistics has been doing a new research in order to identify this Bias, but it has been diversity of opinions from the different individuals investigating them.

In this text there are stated some difficulties in order to elaborate de CPI correctly, which I am going to briefly explain in this essay.

CPI elaborating by the Bureau of Economic Statistics is deflating personal consumption when constructing National Accounts, which leads to Biased measures of real growth and productivity.

The magnitude of these Biases would bring us consequences of 19% increasing in cost living in a 25-year period, but larger Biases could have an even more problematic change in cost of living.

The CPI measures cost of living from a basket of an average consumption products during a certain period. But this index is based on fixed units from an historical nature, which not allows us to take into account the substitution goods when a good gets more expensive or changes in quality.

One of the practical problems is that direct information on based-period quantities is generally not available; instead, the household surveys provide information on expenditures on categories of items.

The first Bias that I’m going to explain is the Commodity substitution Bias. This Bias reflects the failure of a fixed basket of products. If the price of this product increases, then, we move into the substitute product, which is not reflected into the CPI. There are empirical strategies to estimate substitution Bias. If we take into account that equation, then the increase of price of that product brings us to the exact cost of living.

Diewert (1976) The superlative price index allows us to be taken into account the substitute products. Laspeyres grows a 0,2 – 0,25 % faster each year as doesn’t take into account that Bias.

Reindsorf (1993) has been investigating the price scanner of the supermarket in order to study los-level substitution products for a handful of grocery products. These low-level substitution products should also be taken into account.

CPI grows from 1,5 to 2 % faster than it should grow in a year (Reindsorf 1993). This difference is attributes to the Formula Bias. This Bias gives much weight to an increase of price than a decrease of prices.

The Outlet substitution Bias doesn’t compares directly the old and new outlets. The price in new samples is 1,25% lower in the old samples over a two-year period.

There’s also a Bias named quality adjustment bias. There are some products which changes in quality are not really changed so then it’s easy to compare the prices of old and new products. There are also changes in quality very significant, which allows us to compare the new and old product prices. But there are some products which changes in quality are very significant but CPI doesn’t reflect the information in prices.

Another Bias is the new goods bias. The problem is that the CPI brings the new goods to a timely basis. A change in the Point-of-Purchase in order to use more frequently sample replacements would reduce this Bias. The CPI misses the consumer surplus increase, so some products are highly overstated.

In my opinion, the most important Bias is this related to the substitution Bias. It increases grocery prices by 2% a year, which is where we spend a large number of euros every month. By getting this solved we would get ready of 20 euros less by every 100 euros after 10 years on not having solved this problem


Marc Ramon Hernández