This is a repost from earlier this afternoon. Some commentators referred to the fact that a few of my figures were incorrect. This is true. Some figures were wrong. I was in a rush and collating prices from the subtext of Google searches, which resulted in some oversights (as well as accidentally getting an Officeworks price for sugar, instead of Woolworths). I've now cross-referenced each line twice directly from the source sites. I appreciate the effort of prior commentators, but believe that the corrections have not changed my conclusions. Link to the deleted post for full disclosure and transparency: https://www.reddit.com/r/AusFinance/comments/segps3/the_abs_figure_for_the_food_component_of_the_cpi/
First, to preface this post, the reason why I'm writing on this topic is because of the increasingly heated arguments going on here between those that accept the CPI figure provided by the ABS/RBA and those that reject it. The intention of this post isn't to completely shutdown the assertions being made ABS/RBA, but rather to create dialogue.
Anecdotal evidence does not count
Here on Reddit you can find countless posts across every subreddit (whether finance related or not) regarding the perceived increase in food prices. The consensus is that food prices have risen exponentially. Outside of Reddit, you will no doubt find co-workers, friends, and relations talking about the skyrocketing price of food. More broadly, you can also find journalists reporting directly from the ground, across the world (including some regions that don't even have internet) and interviewing people that have reported massive increases in the price of food.
But none of that counts - it's all anecdotal. Like, the Dancing Plague of 1518 in Saxony, we are all victims of a mass psychosis in which we perceive food to be more expensive than it actually is. So let's find a way to overcome this psychosis.
The Basket
Initially I began collating prices for a variety of groceries from archives. I decided this probably wouldn't be the best set of results to post since I would probably be accused of cherry-picking items that supported my hypothesis that food CPI is higher than being reported.
Instead, I looked for independent sources that could give me a hypothetical food basket to use. I found one in the consumer advocacy magazine that I'm sure most of us are familiar with, Choice. In April 2021, Choice conducted a survey of a pretty average grocery basket and collected prices at various Australian supermarkets to see which were cheapest, with products of differing brands also being analysed (e.g., home brand, national brand, etc). Choice conducted this research in previous years, but some archived webcrawlers show that (i) the food basket in 2018 was different, and (ii) weights/brands, etc where not presented. As such, I chose to measure only from April 2021. But my research with my own 'foodbasket' using 2020 prices shows that the results are fairly congruent.
The nice thing about the Choice food basket is that it mostly comprised of products that aren't really seasonal - e.g., bread, chocolate, crackers, frozen peas, canned tuna, tinned tomatoes, etc.
Screenshot of basket comparisons to 2022
I decided to simply take the national brand basket from Woolworths and the budget basket for Coles. I avoided IGA because it's a franchise and Aldi because prices are hard to collect online.
Woolworths 'national brand' basket:
Coles 'budget' basket:
Results
Assuming equal monthly compounding prices, the Coles budget basket's annualised inflation was 9.96%. The Woolworths national brand basket's annualised inflation was 6.12%.
The Woolworths inflation figure would've been in the double digits had it not been for an expensive detergent that was reduced by more than $7. That is likely to have caused a severe underestimate since detergent prices are easy to cut. They have insane margins.
Wait - that's way too high!
I'm already aware some of you are hunched over your keyboards, angrily typing away "you didn't measure the price of every item, this is bullshit - the number can't be that high". Well, well well. Thankfully we have sources other than the ABS or me (the random dumbass redditor).
Interest NZ collects data on grocery prices in New Zealand and Australia. It's hypothetical basket is based on a nominal healthy food basket (using the cheapest price of the food item it can find) from the Good Food Magazine. https://www.interest.co.nz/charts/prices/grocery-prices. From 27 Jan 2021 to 26 Jan 2022, the price of the basket rose from $148.22 to $179.2. This is an annual rate of 20.90%.
"Interest NZ is just some random website" I hear you shout in frustration, shaking your fist at the screen. It's OK, relax, take a breath. I have more sources.
The Food and Agriculture Organization, an agency of the United Nations tasked with responding to food security produces a yearly, global index of food prices. Guess what they found? 28% annual inflation in food prices over the course of 2021.
Annual food CPI of 1.9% is simply not possible
At the risk of sounding overly self-assured, I believe there is a 0% chance that the cost of food only rose by 1.9% in Australia.
How could the ABS be underestimating inflation? My best guess is that it's probably a result of some sort of sampling error. The ABS adduces the hypothetical basket of foods by analysing data from checkouts across Australia. Why does this basket have an inflation rate so much lower than other foods? I posit that it's because people, increasingly financially pressured (e.g., 70% of Sydney renters in rent stress), resort to picking the cheapest products on the shelf. This understates inflation for two reasons (i) the food basket automatically becomes cheaper and (ii) cheaper foods are likely to be highly processed foods that are not as susceptible to inflation because of their naturally high margins. As such, the ABS finds that consumers in Australia are only spending a tiny bit more on food. How surprising? The ABS should not be able to adjust the proportion of foods making up the weighting. It creates sampling bias that leads to a large discrepancy in the figures they are reporting and reality.
On the question of supply chains
I won't say much here other than there's always an excuse as to why inflation is running hot - with pundits refusing to admit that low rates and QE have contributed to inflation. In his latest commentary on the markets, Aswath Damodaran, NYU Professor of Corporate Finance, and one of the greatest minds in value investing, had the following to say about supply chains being blamed as the primary cause of inflation:
Initially, the high inflation numbers were attributed to the speed with the economy was recovering from COVID, and once that excuse fell flat, it was the supply chain that was help responsible. By the end of 2021, it was clear that this bout of inflation was not as transient a phenomenon as some had made it out to be
Make of that what you will. This post isn't a debate on the cause of inflation. That can come later.
For now, I wish to talk about why the ABS' CPI figure for food is so divorced from reality.
First, to preface this post, the reason why I'm writing on this topic is because of the increasingly heated arguments going on here between those that accept the CPI figure provided by the ABS/RBA and those that reject it. The intention of this post isn't to completely shutdown the assertions being made ABS/RBA, but rather to create dialogue.
Anecdotal evidence does not count
Here on Reddit you can find countless posts across every subreddit (whether finance related or not) regarding the perceived increase in food prices. The consensus is that food prices have risen exponentially. Outside of Reddit, you will no doubt find co-workers, friends, and relations talking about the skyrocketing price of food. More broadly, you can also find journalists reporting directly from the ground, across the world (including some regions that don't even have internet) and interviewing people that have reported massive increases in the price of food.
But none of that counts - it's all anecdotal. Like, the Dancing Plague of 1518 in Saxony, we are all victims of a mass psychosis in which we perceive food to be more expensive than it actually is. So let's find a way to overcome this psychosis.
The Basket
Initially I began collating prices for a variety of groceries from archives. I decided this probably wouldn't be the best set of results to post since I would probably be accused of cherry-picking items that supported my hypothesis that food CPI is higher than being reported.
Instead, I looked for independent sources that could give me a hypothetical food basket to use. I found one in the consumer advocacy magazine that I'm sure most of us are familiar with, Choice. In April 2021, Choice conducted a survey of a pretty average grocery basket and collected prices at various Australian supermarkets to see which were cheapest, with products of differing brands also being analysed (e.g., home brand, national brand, etc). Choice conducted this research in previous years, but some archived webcrawlers show that (i) the food basket in 2018 was different, and (ii) weights/brands, etc where not presented. As such, I chose to measure only from April 2021. But my research with my own 'foodbasket' using 2020 prices shows that the results are fairly congruent.
The nice thing about the Choice food basket is that it mostly comprised of products that aren't really seasonal - e.g., bread, chocolate, crackers, frozen peas, canned tuna, tinned tomatoes, etc.
Screenshot of basket comparisons to 2022
I decided to simply take the national brand basket from Woolworths and the budget basket for Coles. I avoided IGA because it's a franchise and Aldi because prices are hard to collect online.
Woolworths 'national brand' basket:
Coles 'budget' basket:
Results
Assuming equal monthly compounding prices, the Coles budget basket's annualised inflation was 9.96%. The Woolworths national brand basket's annualised inflation was 6.12%.
The Woolworths inflation figure would've been in the double digits had it not been for an expensive detergent that was reduced by more than $7. That is likely to have caused a severe underestimate since detergent prices are easy to cut. They have insane margins.
Wait - that's way too high!
I'm already aware some of you are hunched over your keyboards, angrily typing away "you didn't measure the price of every item, this is bullshit - the number can't be that high". Well, well well. Thankfully we have sources other than the ABS or me (the random dumbass redditor).
Interest NZ collects data on grocery prices in New Zealand and Australia. It's hypothetical basket is based on a nominal healthy food basket (using the cheapest price of the food item it can find) from the Good Food Magazine. https://www.interest.co.nz/charts/prices/grocery-prices. From 27 Jan 2021 to 26 Jan 2022, the price of the basket rose from $148.22 to $179.2. This is an annual rate of 20.90%.
"Interest NZ is just some random website" I hear you shout in frustration, shaking your fist at the screen. It's OK, relax, take a breath. I have more sources.
The Food and Agriculture Organization, an agency of the United Nations tasked with responding to food security produces a yearly, global index of food prices. Guess what they found? 28% annual inflation in food prices over the course of 2021.
Annual food CPI of 1.9% is simply not possible
At the risk of sounding overly self-assured, I believe there is a 0% chance that the cost of food only rose by 1.9% in Australia.
How could the ABS be underestimating inflation? My best guess is that it's probably a result of some sort of sampling error. The ABS adduces the hypothetical basket of foods by analysing data from checkouts across Australia. Why does this basket have an inflation rate so much lower than other foods? I posit that it's because people, increasingly financially pressured (e.g., 70% of Sydney renters in rent stress), resort to picking the cheapest products on the shelf. This understates inflation for two reasons (i) the food basket automatically becomes cheaper and (ii) cheaper foods are likely to be highly processed foods that are not as susceptible to inflation because of their naturally high margins. As such, the ABS finds that consumers in Australia are only spending a tiny bit more on food. How surprising? The ABS should not be able to adjust the proportion of foods making up the weighting. It creates sampling bias that leads to a large discrepancy in the figures they are reporting and reality.
On the question of supply chains
I won't say much here other than there's always an excuse as to why inflation is running hot - with pundits refusing to admit that low rates and QE have contributed to inflation. In his latest commentary on the markets, Aswath Damodaran, NYU Professor of Corporate Finance, and one of the greatest minds in value investing, had the following to say about supply chains being blamed as the primary cause of inflation:
Initially, the high inflation numbers were attributed to the speed with the economy was recovering from COVID, and once that excuse fell flat, it was the supply chain that was help responsible. By the end of 2021, it was clear that this bout of inflation was not as transient a phenomenon as some had made it out to be
Make of that what you will. This post isn't a debate on the cause of inflation. That can come later.
For now, I wish to talk about why the ABS' CPI figure for food is so divorced from reality.