After a year that saw considerable attention to the rising cost of groceries as Canadians faced a cost-of-living crisis, Canada’s Food Price Report predicts food prices will continue to increase in 2024 – but by less than last year.
The annual forecast, prepared by the University of Guelph, Dalhousie University, University of British Columbia and the University of Saskatchewan, predicts food prices will rise by 2.5-4.5 per cent – an increase of up to $701.79 for a family of four.
This is less than the increase of 5-7 per cent that was predicted for 2023, but still challenging for many families whose budgets have been stretched by rising costs of housing, utilities and debt.
One of the most striking findings of the report was that Canadians spent less on food in 2023, despite rising prices. This suggests that Canadians may be reducing either the quality or the amount of food that they are buying, perhaps as a result of other rising costs.
“Last year was a challenging year for consumers, with rising food prices and increasing levels of food insecurity impacted by climate change, geopolitics, and other factors,” said Dr. Evan Fraser, Director of Arrell Food Institute and University of Guelph Lead on the Food Price Report.
“The Food Price Report predicts we will see further increases to food prices, though thankfully not at the same level as last year. Nevertheless, this may put additional strain on consumers and exacerbate levels of food insecurity.”
Why are food prices rising?
Canadians may wonder what is behind the steep rise in food costs, with some suggesting that price gouging, profiteering or carbon taxes may play a role. The Food Price Report found little evidence for these explanations. Although grocery profits did rise since the pandemic began, this was largely due to inflation; indeed, most price markups occurred in 2020, but were negative or zero in 2022.
The impact of carbon taxes, meanwhile, are complex and difficult to determine. “It would be misleading to assert that carbon pricing has a direct and straightforward impact on retail food prices, and it would be equally misleading to claim otherwise,” says the report.
Meanwhile, climate change and geopolitics are expected to contribute to food price fluctuations, and the rising cost of food, high interest rates and other increasing costs are expected to put additional strain on Canadians’ wallets.
Predictions for 2023 were largely accurate
Predictions from 2023 saw food prices within the expected range, except for bakery items, which cost more than expected, and dairy, which cost less. Amongst G7 countries, Canada had the third lowest food inflation rate.
However, Canadians are spending less than expected on food – in 2023, Canadians reduced the amount they spent on food and beverage retail by 3.26%. This resulted in lower-than-expected food expenditures overall in 2023.
University of Guelph expertise improving predictions
Expertise from the U of G’s Machine Learning Research Group and the Vector Institute for Artificial Intelligence is crucial to ensure the accuracy of predictions in the report. Dr. Graham Taylor and colleagues Kristina Kupferschmidt, Dr. Ethan Jackson, Sara El-Shawa and Cody Kupferschmidt, using state-of-the-art machine learning techniques, are working to improve the analytic models that are the backbone of the Food Price Report.
This year saw the team taking the lead on a new approach. With the advice of subject matter experts, they identified a series of factors that could impact food prices – including information on climate, economics, and geopolitics.
“Last year saw the return of El Niño, which can have major impacts on precipitation and temperatures in North America,” said Cody Kupferschmidt. “By including climatic factors such as El Niño, drought, and water availability in our model, we were able improve predictions for several categories including fruit and vegetables.” Including these factors in the model helped the team to understand how these factors impact food prices and more accurately predict prices in the 2024 Food Price Report.
Other new strategies involved the use of transformer-based machine learning models – similar to the underlying technology used by Chat GPT – to improve predictions, and using ensembles, or the combination of multiple models rather than one model, to ensure accuracy. This innovative approach improved the accuracy of predictions for various categories of food.
Kristina Kupferschmidt emphasized the significance of this approach in improving predictions. “Transformers are an amazing technology, but in order for them to perform well, they require large amounts of high-quality data. We found that by incorporating data from the key factors suggested by food economists, we were able to leverage these sophisticated models for more accurate forecasts.”
Support for this year’s report comes from Arrell Food Institute at U of G, the Vector Institute and Dalhousie’s Agri-Food Analytics Lab. Other U of G authors and advisors involved in the report were Dr. Maria Corradini, Dr. Jess Haines, and Paul Uys.