Transcript of the chat session Consumer Price Index: Basket update, which occurred on Friday, February 27, 2015 from 1:30 to 2:30 p.m. EST

Note: This was a bilingual chat session, which means that the participants were able to submit their questions in English or French. Statistics Canada respects the Official Languages Act and is committed to ensuring that information products of equal quality are available in both English and French. For that reason, all the questions and answers have been translated in the other official language.

 Moderator at 13:30:00
Welcome everyone! This is a bilingual chat session, which means that you can submit your questions in English or French. Our experts will respond in a timely manner and in the official language in which the question was asked.

 Kira Judge at 13:32:00
Hi All
do I need to adjust my previous months data to make comparable with January data?

 Kyle de March at 13:36:33
G'day Kira,
Great question.
No, you do not need to adjust your previous months' data to make them comparable. Each time we do a basket update, we chain link the indexes to ensure a continuous time series.
Thank you!
Kyle

 sdimino at 13:33:00
Hi Silvana Dimino here from JPMorgan Economics dept. Can you give me an idea of which weights went up the most and which went down the most

 Mathieu Lequain at 13:42:05
Hi sdimino
Weights are available at (incorrect link provided)
Biggest decreases:
Canned and other preserved fish
Flour and flour-based mixes
Leasing of passenger vehicles
Other health care goods
Fuel, parts and accessories for recreational vehicles
Audio equipment
Rental of digital media
Newspapers
Magazines and periodicals
Other reading material (excluding textbooks)
For increases:
Oranges
Tomatoes
Taxi and other local and commuter transportation services
Sporting and exercise equipment
Multipurpose digital devices
Spectator entertainment (excluding video and audio subscription services)
Thank you for your question.

 ChristineL at 13:33:00
What is the purpose of a basket update?

 Amanda Wright at 13:43:15
The main purpose of a basket update is to ensure that the weights used in the calculation of the CPI are representative of consumer spending habits. The CPI uses a "fixed" basket that becomes out of date over time because consumers tend to change their buying habits as relative prices change. Updating the basket frequently helps reduce any bias in the index that could result from the use of a "fixed" basket.

 Dawn Chipps at 13:34:00
Can you explain the "linking" that is used to maintain continuity with the recent change in the basket weighting?

 Bruno Morin at 13:43:40
Hello,
The linking is well explained in the Canadian CPI reference paper in Chapter 8 (8.13 to 8.21) Chapter 8 – Weights and Basket Updates.
Published consumer price indices are calculated as a chain of fixed-basket indices. This means that a sequence of fixed-basket indices have been chained together to create a continuous time series. This type of chaining is not to be confused with the calculation of monthly chained indices but rather refers to the process of chaining indices across baskets. This is necessary to avoid having breaks in an index when a basket update is performed.

 bounajm at 13:37:00
Hi there. Has Statistics Canada ever attempted to measure the impact of switching from a four-year to a two-year basket update? Presumably, it will have slowed inflation down a bit, but it would be nice to know if that was the case, and by how much. Thanks.

 Kyle de March at 13:46:33
Hi bounajm,
This basket, using expenditures derived from the 2013 Survey of Household Spending, was updated after two years. The introduction of the 2011 basket weights with the February 2013 CPI marked the first time the basket was updated after two years. Previously, it had been done every four years, generally.
We are currently working on an analytical series that will analyze the impact of the frequency of the basket update on the measure of the CPI.
Thanks for your question!
Kyle

 pierre-guy sylvestre at 13:45:00
Hi, I am Pierre-Guy Sylvestre from the Canadian Union of Public Employees. I would like to know: Do you have an estimate of how much this change will reduce the upward bias in the CPI?

 Mathieu Lequain at 13:48:56
Hi Pierre-Guy
We are currently working on an analytical series that will analyse the impact of the frequency of the basket on the measure of the CPI.
Thank you

 TUSHART at 13:38:00
CPI Excluding energy seems low compared to US or other Nation Please can shed some light why this is so?

 Amanda Wright at 13:53:41
Making direct comparisons between the CPIs of different countries can be challenging because of several factors:
- not all products are comparable and could be influenced by unique market factors
- most services are priced to the local market and therefore may not be comparable across regions
- CPI methodologies can differ, notably in the area of shelter costs.
However, sometimes comparisons can be made. For example the decline in energy prices, particularly for gasoline, is something many countries are currently experiencing. This is largely impacted by the fall in oil prices. Excluding energy, the Canadian CPI increased 2.3% year over year in January. Factors contributing to this rise were higher food prices, notably for fruits, vegetables and meat.

 Kira Judge at 13:46:00
The link you provided for sdimino is comming up as an error page is there other way of getting the same info please

 Mathieu Lequain at 13:54:57
Sorry my mistake.
CANSIM number Table 326-0031
This link should work: Basket Weights of the Consumer Price Index
Thank you

 TUSHART at 13:46:00
Importance of food is is 16% and the Price Index is up from last year 133 to 135, While transportation importance is 19% and Price index has dropped from 129 to 122 has weights caused it?

 Bruno Morin at 13:58:21
No. The price index movements are driven by monthly price changes, while the importance in the All-items CPI (weights) is derived from expenditure weights obtained from the Survey of Household Spending.

 pierre-guy sylvestre at 13:51:00
How much upward bias do you estimate exists in the CPI due to the “substitution effect”, both of products and outlets , and of other factors?

 Amanda Wright at 14:04:00
It is well known that there is bias in the CPI. The bias results from using a "fixed" basket of consumer spending when we know that consumers are constantly adjusting their buying habits in response to changing relative prices, and the introduction of new goods or outlets. Updating the CPI basket frequently helps minimize the bias over time.
The size of the commodity-substitution bias can be measured by comparing the official CPI series with a measure of the cost of living (COLI). Using a retrospective Fisher index for the COLI, we can estimate the size of the commodity-substitution bias over time by calculating what could have been the measure of the CPI if Statistics Canada was to calculate a COLI.
The Bank of Canada has published papers on the biases in the CPI, including the product-substitution bias. Statistics Canada is also working on research to study the impact of more frequent basket updates.

 SmallFlockers at 13:57:00
It is well known that consumers will alter their buying habits due to price fluctuations in the marketplace and their earnings. Health Canada recently reported that 7.6% of Canadian families can't afford the food they need to feed their families. In Nunavut, 75% of pre-school children do not eat on any particular day because there is no food in the house, and some families are "shopping" for their food in the local garbage dump. If food and other commodities continue to rapidly inflate so that the food affordability inflicts more and more damage to Canadian families, so that some day in the future, a majority of Canadians can no longer buy groceries, but 99.9% of Canadians rely upon garbage dumps (and/or free earthworms, grubs, and grass growing in fields around their home) to feed themselves, and the CPI basket is then updated according to the 0.01% rich consumers who still shop in grocery stores, does StatsCan have any plans to change the current methodology of updating the CPI and the basket definition to reflect the buying habits of Canadians on that fateful day? Are there any advantages for Canadians and the Government of Canada to start using that future method for CPI basket adjustment today?

 Kyle de March at 14:07:54
Hi SmallFlockers,
While it is possible to construct consumer price indexes for specific population groups (as was done for the studies on seniors and income groups), the official CPI is produced to represent price change experienced by the average household. Families and individuals living in urban and rural private households in Canada are the target population for the CPI.
The CPI expenditure weights are estimated primarily from data taken from Statistics Canada’s Survey of Household Spending, which is conducted annually and yields statistical estimates of household expenditure by product class and region. Each household contributes to the total weight of a product group proportionally to its respective spending.
This weighting method is the most commonly used amongst the international community of statistical agencies producing consumer price indexes. No plans are in place to deviate from the current method of weighting; however, continual research at Statistics Canada ensures that informed decisions in regard to methodological matters are made when required.
Thanks for your question,
Kyle

 pierre-guy sylvestre at 14:01:00
Thank you Mathieu Lequain for the answer. I have two last questions: What other types of bias exist in the CPI and what other changes have or are being made to reduce them? How much of this has been reduced by different changes to methodology of calculating the CPI? Do you have those calculations by province?

 Mathieu Lequain at 14:12:40
Thank you Pierre-Guy.
In the case of the CPI, bias can occur for several reasons: when one product is replaced with another and an associated quality adjustment has to be made (quality adjustment bias). Potential bias is associated with new product introductions and outlet substitutions. Bias might also be inherent in some editing procedures.
The Bank of Canada published several papers referring to the different types of bias, at the Canada level only.
Merci. Thank you

 thestatman at 14:00:00
Hi, does the update worsen the inflation faced by seniors?

 Amanda Wright at 14:14:32
While it is possible to construct consumer price indexes for specific population groups (as was done for the studies in the past for seniors and income groups), the official CPI is produced to represent price change experienced by the average household. All families and individuals living in urban and rural private households in Canada are the target population for the CPI.
Updating the CPI basket helps ensure that the CPI basket remains representative of all consumer spending in Canada.
The inflation experience of any particular household or population group can change over time. It depends on the price trends of the products they buy more frequently. Sometimes the items they buy more frequently can increase in price while at other times they may fall.

 bounajm at 14:05:00
Hi again, thanks for answering my first question. I look forward to the analytical series that will analyze the impact of the frequency of the basket on the measure of the CPI.
I hope it's okay to ask a question that is not exactly related to the CPI basket. Does Statistics Canada make adjustments to prices based on quality improvements (or deterioration)? I believe that is the case, but I could not find any documentation on the website that goes into detail explaining how these adjustments are done for each good and service. I would also like to know if these adjustment methodologies similar across different countries (US, Europe)? thanks again

 Kyle de March at 14:16:35
Bounajm, you're on a roll.
The Canadian Consumer Price Index aims to measure pure price change, thus excluding price changes that are due to differences in the quality of products bought by consumers. It achieves this mostly through the matched-model approach which tracks unchanging products in the same outlets, thus holding all variables constant except for the month when prices are observed.
Statistics Canada uses internationally accepted methods for quality change and adjustment.
For more information on quality change on adjustment, please see Chapter 7 of The Canadian Consumer Price Index Reference Paper. Have a nice day!
Kyle

 Kira Judge at 14:14:00
THank you guys for very detailed answers. One more question from me: I appologize if it is off topic but would like to know when the IPI tables will be updated with January data?

 Moderator at 14:17:30
Kira Judge, thank you for your question. We will follow up with you by email in the following business days.

 SmallFlockers at 14:14:00
Kyle, thank-you for your answer, but it doesn't seem to be on point. The CPI is all about "purchases". However, there is the substitution effect of people forced to non-purchases (ie. Food Banks, charity, soup kitchens, garbage dumps, dumpster diving, etc.). As more and more people are forced to non-purchases, how will that affect the CPI basket, likely skewing and biasing it more and more. Does StatsCan see that as a problem? Are there other measurements that will show this growing bias? Is there an adjustment that should be made to the CPI basket, or a note warning of this bias?

 Kyle de March at 14:25:39
To calculate the CPI, we need three elements: an expenditure to give relative importance to each price movement; we need to be able to estimate the transaction price paid by a consumer; and we need to be able to observe the characteristics of the products purchased to estimate a pure price movement that removes the effect of the quality change.
Thanks for your follow-up question. Have a great day!
Kyle (Credit to Mathieu Lequain for this response!)

 thestatman at 14:24:21
Thank you Amanda. If Inflation for seniors is ever updates , can I be put on the distribution list?

 pierre-guy sylvestre at 14:24:06
Thanks Amanda Wright for the answer. You spoke about the COLI: Are there other changes planned to make the CPI more reflective of changes in the cost of living, with adjustments to depreciation rates or etc.

 CHANSUE at 13:47:00
Could you explain a bit about the difference in "Weight at basket link month prices" and "Weight at basket reference period prices"? The "Weight at basket link month prices" is the one normally used?
Thanks

 Kyle de March at 14:26:01
Hi CHANSUE,
Good question. I imagine you see these two options available on our CANSIM page.
The weights used for the CPI basket weights are coming from the 2013 Survey of Household Spending. When people report their expenditures, they report the value of the goods and services consumed at the prices from that year. The link month weights use this information but value their expenditures not at the reference period prices, but the prevaling prices in the link month - in this case, December, 2014.
Thanks for your question,
Kyle

 Kira Judge at 13:40:00
I want to know answer to Tushart question please:
CPI Excluding energy seems low compared to US or other Nation Please can shed some light why this is so?

 Amanda Wright at 13:53:41
Making direct comparisons between the CPIs of different countries can be challenging because of several factors:
- not all products are comparable and could be influenced by unique market factors
- most services are priced to the local market and therefore may not be comparable across regions
- CPI methodologies can differ, notably in the area of shelter costs.
However, sometimes comparisons can be made. For example the decline in energy prices, particularly for gasoline, is something many countries are currently experiencing. This is largely impacted by the fall in oil prices. Excluding energy, the Canadian CPI increased 2.3% year over year in January. Factors contributing to this rise were higher food prices, notably for fruits, vegetables and meat. I hope that answers your question.

 Kira Judge at 13:39:36
thanks Kyle for your answer and have a wonderfull weekend
Kira

 Dawn Chipps at 14:28:56
Thanks Bruno for your response.

 tobysanger at 14:21:00
How much upward bias is the CPI estimated to have, as a result of the substitution effect and other factors and what other changes are or have been implemented to reduce it?

 Bruno Morin at 14:30:22
The Bank of Canada has published several papers on these topics.
There is a consensus among CPI experts that updating the basket more frequently could in the long run reduce the substitution bias. In terms of other biases due to the introduction of new products or new outlets, Statistics Canada manages its sample in a dynamic manner to reflect very quickly those changes in the marketplace.

 LINDALI at 14:20:00
Hi, Energy price plays a critical role in our lives and business, I believe excluding such an impactful factor in the calculation would mild the downward trend, well the question is whether it's misleading to the market...?

 Amanda Wright at 14:32:27
Absolutely, energy is a critical element in the economy and in our lives and it is included the All-items CPI. All of the products households buy are included in the CPI.
With the most recent CPI basket update energy products represent 7.79% of the total basket of consumer spending.
Statistics Canada does calculate and publish special aggregate indexes, many of which exclude certain products. One special aggregate is the All-items excuding energy. This and other special aggregate series are used by businesses, policy analysts, economists and households to gauge the contribution of the remaining products to overall price change. For instance, on a year-over-year basis the CPI excluding energy rose 2.3% while the All-items CPI rose 1.0%. This indicates that energy prices were exerting downward pressure on the CPI in the 12-months to January, becuase if excluded inflation would have come in higher than what the official CPI did.

 thestatman at 14:07:00
I think i was not clear..sorry, allow me to rephrase: does the new basket reduces the effectiveness of the CPI as a measure of inflation faced by Canadian seniors?

 Moderator at 14:32:59
Please see Amanda's response:

 Amanda Wright at 14:14:32
While it is possible to construct consumer price indexes for specific population groups (as was done for the studies in the past for seniors and income groups), the official CPI is produced to represent price change experienced by the average household. All families and individuals living in urban and rural private households in Canada are the target population for the CPI.
Updating the CPI basket helps ensure that the CPI basket remains representative of all consumer spending in Canada.
The inflation experience of any particular household or population group can change over time. It depends on the price trends of the products they buy more frequently. Sometimes the items they buy more frequently can increase in price while at other times they may fall.

 Moderator at 14:30:00
The chat session is now over. Thank you for your questions and comments! If our experts did not have a chance to respond to your question, we will follow-up with you by email in the next few business days. The full transcript of this chat session will be made available on our website shortly. Have a great day!

Date modified: