Did the BLS "Consumer Price Index New Release" for July contain a major error in gasoline prices?

Ask an Economist

I have a question - regarding the BLS "Consumer Price Index New Release" for July:


Did they make a mistake - in their report of gasoline prices (or have I've misunderstood something)?

They reported a July unadjusted percent change - for gasoline (all types) = .2%
(see link below)


I follow gasoline prices - as reported by the U.S. Energy Information Agency (EIA), and by others like the AAA. The EIA and AAA reports for July - were consistent. Their reports showed a July increase of - approximately 20 cents / gallon.

If the EIA & AAA reports are correct, the July unadjusted percent change (for all types of gasoline) was: .20 / 3.66 = 5.47%
not: .2%

Such a sizable difference - would significantly impact the reported CPI, and other averages.

Here's the relevant EIA data:

At the beginning of July, EIA reports for U.S. gasoline price [all grades all formulations] were:

On 06/26/2023 - They reported all grades all formulations = 3.685
On 07/03/2023 - They reported all grades all formulations = 3.643

An estimate for 07/01/2023 - using the 06/26 and 07/03 prices - was 3.66
On 07/31/2023 - EIA reported [all grades all formulations] gasoline price = 3.869
So their prices indicate a July increase of 20 cents / gallon.

Note: The same exercise was done using the EIA Regular Gasoline Prices - which also showed an increase of about 20 cents / gallon.

Here's the relevant AAA report (Aug 3):

A key statement in their Aug 3rd report says,
"Today’s national average of $3.82 is 29 cents more than a month ago"

Any thoughts you might have regarding my question of the July BLS report - is greatly appreciated…

The EIA website - also has History spreadsheet (with comprehensive data)



The following is a snapshot of the EIA report regarding Regular Gasoline Prices on 07/03:
price was 3.527, and change was -0.044 from week ago - from which I noticed about the same 20 cents / gallon increase in July, for regular gasoline.


The BLS reports Consumer Price Indices which are different from nominal prices reported by other sources mentioned in the question. The BLS index measures price change from a designated reference date (the reference base for most of the items is 1982-84 equals 100). Prices are collected each month in 75 urban areas across the country from about 6,000 housing units and approximately 22,000 retail establishments. In calculating the index, price changes for the various items in each location are aggregated using weights, which represent their importance in the spending of the appropriate population group. Local data are then combined to obtain a U.S. city average (Please see the technical note at https://www.bls.gov/news.release/cpi.htm). It is important to note that prices used to compute the CPI are collected during the entire month. CPI data is published monthly, with the index value representing an estimate of the price level for the month as a whole, rather than a specific date (Please see question number 11 at https://www.bls.gov/cpi/questions-and-answers.htm). As per the data release, the unadjusted indices for Gasoline were 318.071 and 316.258 in July and June, respectively, representing an increase of 0.6 percent.  

The EIA data, however, are nominal prices in dollar collected on every Monday on a weekly basis from approximately 1,000 retail gasoline outlets. The reported price is the cash pump price paid by a consumer as of 8:00 am Monday. They are then used to calculate volume-weighted average gasoline price estimates at the national level (Please see the explanatory notes at https://www.eia.gov/dnav/pet/TblDefs/pet_pri_gnd_tbldef2.asp). The two data series, therefore, are not strictly comparable, but the percentage changes should not be too different, as pointed out in the question.

The problem seems to be the following: data corresponding to 07/31 and 06/26 reported by EIA are of those specific Mondays and do not represent the prices of the entire month, while CPI indices represent the price level for the month as a whole, rather than a specific date, as mentioned earlier.  To find a comparable dataset, one may download the monthly series from EIA (link: https://www.eia.gov/dnav/pet/hist/LeafHandler.ashx?n=pet&s=emm_epm0_pte_nus_dpg&f=m). As per this series (this is an average of the weekly collected data), July and June prices were  $3.712 and $3.684 per gallon, implying a month-on-month increase of 0.76 percent, fairly close to the BLS report of 0.6 percent (we have to consider the unadjusted data as EIA is unadjusted). The slight difference observed here could be because of their differing data collection methods and/or varying weights used in aggregation.  

Answered by:
Ph. D. Student
Last updated on September 3, 2023